Humanoid Robots Wiki wikidb https://humanoids.wiki/w/Main_Page MediaWiki 1.31.0 first-letter Media Special Talk User User talk Humanoid Robots Wiki Humanoid Robots Wiki talk File File talk MediaWiki MediaWiki talk Template Template talk Help Help talk Category Category talk Gadget Gadget talk Gadget definition Gadget definition talk 0 1 1 2018-07-06T01:27:54Z MediaWiki default 0 wikitext text/x-wiki <strong>MediaWiki has been installed.</strong> Consult the [https://www.mediawiki.org/wiki/Special:MyLanguage/Help:Contents User's Guide] for information on using the wiki software. == Getting started == * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Configuration_settings Configuration settings list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:FAQ MediaWiki FAQ] * [https://lists.wikimedia.org/mailman/listinfo/mediawiki-announce MediaWiki release mailing list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation#Translation_resources Localise MediaWiki for your language] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Combating_spam Learn how to combat spam on your wiki] 5702e4d5fd9173246331a889294caf01a3ad3706 2 1 2018-07-07T00:15:06Z 199.87.196.213 0 wikitext text/x-wiki <strong>MediaWiki has been installed.</strong> <syntaxhighlight lang=bash> #!/bin/bash </syntaxhighlight> Consult the [https://www.mediawiki.org/wiki/Special:MyLanguage/Help:Contents User's Guide] for information on using the wiki software. == Getting started == * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Configuration_settings Configuration settings list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:FAQ MediaWiki FAQ] * [https://lists.wikimedia.org/mailman/listinfo/mediawiki-announce MediaWiki release mailing list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation#Translation_resources Localise MediaWiki for your language] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Combating_spam Learn how to combat spam on your wiki] 4a8a61f37b0f1f2f01d7826a5c6559917443c54b 3 2 2018-07-07T00:15:35Z 199.87.196.213 0 wikitext text/x-wiki <strong>MediaWiki has been installed.</strong> <syntaxhighlight lang=bash> #Test of Syntax Highlight </syntaxhighlight> Consult the [https://www.mediawiki.org/wiki/Special:MyLanguage/Help:Contents User's Guide] for information on using the wiki software. == Getting started == * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Configuration_settings Configuration settings list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:FAQ MediaWiki FAQ] * [https://lists.wikimedia.org/mailman/listinfo/mediawiki-announce MediaWiki release mailing list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation#Translation_resources Localise MediaWiki for your language] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Combating_spam Learn how to combat spam on your wiki] 4dac9928604cac8ef81cd55d34940567098a4806 4 3 2018-07-07T00:15:53Z 199.87.196.213 0 wikitext text/x-wiki <strong>MediaWiki has been installed.</strong> <syntaxhighlight lang=bash> # Test of Syntax Highlight </syntaxhighlight> Consult the [https://www.mediawiki.org/wiki/Special:MyLanguage/Help:Contents User's Guide] for information on using the wiki software. == Getting started == * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Configuration_settings Configuration settings list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:FAQ MediaWiki FAQ] * [https://lists.wikimedia.org/mailman/listinfo/mediawiki-announce MediaWiki release mailing list] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Localisation#Translation_resources Localise MediaWiki for your language] * [https://www.mediawiki.org/wiki/Special:MyLanguage/Manual:Combating_spam Learn how to combat spam on your wiki] c2cbf3dc7cdb438197b1cc1afdb56ce496e4ed42 5 4 2024-04-23T20:12:59Z Ben 2 Home page wikitext text/x-wiki <h1>Humanoid Robot Wiki</h1> Welcome to our wiki! 836a951866be43ca2fc601ebdbb4800916b8bce7 6 5 2024-04-23T20:13:10Z Ben 2 wikitext text/x-wiki Welcome to our wiki! 6f8e6105ab29f72f6943ccdff9a27f733ace71bc 10 6 2024-04-23T23:24:03Z Admin 1 wikitext text/x-wiki Welcome to the humanoid robots wiki! 2201b9727474e5af9b6abd848e8b0729efa1823d 11 10 2024-04-23T23:35:16Z MattFreed 3 Add Hardware Link wikitext text/x-wiki Welcome to the humanoid robots wiki! <big>Navigation:</big> * [[Hardware]] 8541f5b0c47db499dec9f8237202bd9cd7bc411a 16 11 2024-04-24T01:45:50Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! <big>Navigation:</big> * [[Hardware]] bff45cf8138e12849f826ceffc39ca4c66a96aaa 20 16 2024-04-24T01:48:18Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! === Companies === - [[Unitree]] - [[K-Scale Labs]] ad53fe969f480bd8d8b8b4cb8b406054c7148298 22 20 2024-04-24T01:49:13Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! === Companies === * [[Unitree]] * [[K-Scale Labs]] 7da9dc5164c0448c55e025ca6f83cb31491d804b 23 22 2024-04-24T01:51:38Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! === Getting Started === This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === * [[Tesla]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Unitree]] * [[K-Scale Labs]] ee4e3c3d12fa4cce356f15b9ae54b4d0215f20a6 29 23 2024-04-24T01:55:01Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === * [[Tesla]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Unitree]] * [[K-Scale Labs]] b74711193adf859de54225190e3e126c3eb5118b 30 29 2024-04-24T01:57:12Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === ==== Robots ==== * [[Tesla]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Unitree]] * [[K-Scale Labs]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] 042a104263722d2954a02d4a4fc566c1d7934a69 33 30 2024-04-24T01:59:45Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === ==== Robots ==== * [[Tesla]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Unitree]] * [[K-Scale Labs]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] 56cfd94b07f65246df1dafa198025aef6990771d 49 33 2024-04-24T02:07:54Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === ==== Robots ==== * [[Tesla]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Fourier Intelligence]] * [[Unitree]] * [[K-Scale Labs]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] 4441a5e2a290521ca0d33bac770b52bac39385a7 50 49 2024-04-24T02:08:51Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === ==== Robots ==== * [[Tesla]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Fourier Intelligence]] * [[AGIBot]] * [[Unitree]] * [[UBTech]] * [[Boston Dynamics]] * [[Apptronik]] * [[K-Scale Labs]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] 315a37b849144d01a767d17d33bbeb7eead0ed77 0 2 7 2024-04-23T21:33:41Z MattFreed 3 Initial Post wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for K-Scale labs humanoid robot, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robotics]] [[Category:Hardware Components]] 50e6a6d41a9f25139d6fb056099ecaf15449fc9c 8 7 2024-04-23T22:11:29Z MattFreed 3 Actuator Information wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for K-Scale labs humanoid robot, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators The K-Scale Labs robot uses Quasi-direct drive pancake motors based off the MIT Cheetah open-source actuators. These actuators and developed, manufactured, and sold by MyActuator. The K-Scale Labs robot consists of: * 10x MyActuator RMD-X4-H * 6x MyActuator RMD-X6 * 9x MyActuator RMD-X8 * 4x MyActuator RMX-X10 == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robotics]] [[Category:Hardware Components]] b6b03099a16463e8c68216e1c3c7d7f1130d051c 9 8 2024-04-23T22:21:19Z MattFreed 3 spelling wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for K-Scale labs humanoid robot, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators The K-Scale Labs robot uses Quasi-direct drive pancake motors based off the MIT Cheetah open-source actuators. These actuators and developed, manufactured, and sold by MyActuator. The K-Scale Labs robot consists of: * 10x MyActuator RMD-X4-H * 6x MyActuator RMD-X6 * 9x MyActuator RMD-X8 * 4x MyActuator RMX-X-10 == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robotics]] [[Category:Hardware Components]] 55a15c51657f80bdbc3553032d7e0d7f9a0beac9 12 9 2024-04-23T23:38:37Z MattFreed 3 spacing wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for K-Scale labs humanoid robot, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators The K-Scale Labs robot uses Quasi-direct drive pancake motors based off the MIT Cheetah open-source actuators. These actuators and developed, manufactured, and sold by MyActuator. The K-Scale Labs robot consists of: * 10x MyActuator RMD-X4-H * 6x MyActuator RMD-X6 * 9x MyActuator RMD-X8 * 4x MyActuator RMX-X-10 == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robotics]] [[Category:Hardware Components]] 856fdd819e1695b5bed76b440123c74bded7faf4 13 12 2024-04-24T01:31:54Z Admin 1 wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators Many humanoid robots use Quasi-direct drive pancake motors based off the MIT Cheetah open-source actuators. These actuators and developed, manufactured, and sold by MyActuator. The K-Scale Labs robot consists of: * 10x MyActuator RMD-X4-H * 6x MyActuator RMD-X6 * 9x MyActuator RMD-X8 * 4x MyActuator RMX-X-10 == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robotics]] [[Category:Hardware Components]] 676b51065476d9febd5528276531169818feb1ef 14 13 2024-04-24T01:44:51Z MattFreed 3 remove kscale direct information for now wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robotics]] [[Category:Hardware Components]] 0de5719c371084d0f179818e8070e81b74dd382a 17 14 2024-04-24T01:46:20Z Ben 2 Ben moved page [[Hardware]] to [[K-Scale Labs Hardware]]: It is K-Scale Labs specific wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robotics]] [[Category:Hardware Components]] 0de5719c371084d0f179818e8070e81b74dd382a 45 17 2024-04-24T02:03:51Z Ben 2 wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robots]] 85133efd7e50150f4695e1b8df3aa48d96d46d4a 46 45 2024-04-24T02:04:01Z Ben 2 Ben moved page [[K-Scale Labs Hardware]] to [[Stompy]] wikitext text/x-wiki = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robots]] 85133efd7e50150f4695e1b8df3aa48d96d46d4a 0 3 15 2024-04-24T01:45:18Z Ben 2 Created page with "It is available for purchase [https://shop.unitree.com/products/unitree-h1 here]." wikitext text/x-wiki It is available for purchase [https://shop.unitree.com/products/unitree-h1 here]. 900095efde90381d578981a4184f54e668da07cf 0 4 18 2024-04-24T01:46:20Z Ben 2 Ben moved page [[Hardware]] to [[K-Scale Labs Hardware]]: It is K-Scale Labs specific wikitext text/x-wiki #REDIRECT [[K-Scale Labs Hardware]] 53f4d1ebab3f12033dccb67237c2d445b0fecb21 0 5 19 2024-04-24T01:47:48Z Ben 2 Created page with "K-Scale Labs is building an open-source humanoid robot called Stompy. Their website is [https://kscale.dev/ here]." wikitext text/x-wiki K-Scale Labs is building an open-source humanoid robot called Stompy. Their website is [https://kscale.dev/ here]. 40e0942e7db5e0ffc0af846735224cba3dedbce6 28 19 2024-04-24T01:54:51Z Ben 2 wikitext text/x-wiki [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called Stompy. d0533f3cfb47d26fa8f05607edc876d2ff943b0c 35 28 2024-04-24T02:00:37Z Ben 2 wikitext text/x-wiki [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[K-Scale Labs Hardware|Stompy]]. 31bc672d493af7c0dca35c856987fb252ab85c46 36 35 2024-04-24T02:01:26Z Ben 2 wikitext text/x-wiki [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[K-Scale Labs Hardware|Stompy]]. [[Category:Companies]] deae7c6fe0f7d7e5c7fd765fef883a4723a77ea5 48 36 2024-04-24T02:04:15Z Ben 2 wikitext text/x-wiki [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. [[Category:Companies]] d180a3ab7af23ea495ff3e0923e9916efdda6a84 0 6 21 2024-04-24T01:49:03Z Ben 2 Created page with "Unitree is a company based out of China which has built a number of different types of robots. === Robots === * [[Unitree H1]]" wikitext text/x-wiki Unitree is a company based out of China which has built a number of different types of robots. === Robots === * [[Unitree H1]] ab34503c057ce402cafbe936b3b66b1cc1663672 37 21 2024-04-24T02:01:37Z Ben 2 wikitext text/x-wiki Unitree is a company based out of China which has built a number of different types of robots. === Robots === * [[Unitree H1]] [[Category:Companies]] c5679d0cacd2419150766e0bd1dd696257e2a543 0 7 24 2024-04-24T01:51:50Z Ben 2 Created page with "Tesla is building a humanoid robot called Optimus." wikitext text/x-wiki Tesla is building a humanoid robot called Optimus. 47566dbb56714ed62fadd552286ecdf5cd1461ed 41 24 2024-04-24T02:02:03Z Ben 2 wikitext text/x-wiki Tesla is building a humanoid robot called Optimus. [[Category:Companies]] b41b0bd49949661e997e6ba50f698a3fa688aaa1 0 8 25 2024-04-24T01:52:31Z Ben 2 Created page with "Agility has built several robots. Their humanoid robot is called Digit." wikitext text/x-wiki Agility has built several robots. Their humanoid robot is called Digit. b7e714a16ee5fb0e2fb64cf4fabf27734e6fed8d 40 25 2024-04-24T02:01:58Z Ben 2 wikitext text/x-wiki Agility has built several robots. Their humanoid robot is called Digit. [[Category:Companies]] a2902ca85db51df97724bd28f08f427a6e25d0ae 0 9 26 2024-04-24T01:53:25Z Ben 2 Created page with "[https://sanctuary.ai/ Sanctuary AI] is a humanoid robot company. Their robot is called Phoenix." wikitext text/x-wiki [https://sanctuary.ai/ Sanctuary AI] is a humanoid robot company. Their robot is called Phoenix. 9e4a36100b2f4b58649f89c323f0d5a3caf47e36 39 26 2024-04-24T02:01:51Z Ben 2 wikitext text/x-wiki [https://sanctuary.ai/ Sanctuary AI] is a humanoid robot company. Their robot is called Phoenix. [[Category:Companies]] e8604c484200d8b58bd8921b26df01fa6974d4a6 0 10 27 2024-04-24T01:54:29Z Ben 2 Created page with "[https://www.1x.tech/ 1X] (formerly known as Halodi) is the best humanoid robot company. They have two robots: EVE and NEO. EVE is a wheeled robot while NEO has legs." wikitext text/x-wiki [https://www.1x.tech/ 1X] (formerly known as Halodi) is the best humanoid robot company. They have two robots: EVE and NEO. EVE is a wheeled robot while NEO has legs. f536513b93f4310ecdf1505345534b7523e06a07 38 27 2024-04-24T02:01:45Z Ben 2 wikitext text/x-wiki [https://www.1x.tech/ 1X] (formerly known as Halodi) is the best humanoid robot company. They have two robots: EVE and NEO. EVE is a wheeled robot while NEO has legs. [[Category:Companies]] 2206882797e51be1b841dde3d4f754333fa1c205 0 11 31 2024-04-24T01:58:05Z Ben 2 Created page with "Skild is a stealth foundation model startup started by two faculty members from Carnegie Mellon University. === Articles === * [https://www.theinformation.com/articles/ventu..." wikitext text/x-wiki Skild is a stealth foundation model startup started by two faculty members from Carnegie Mellon University. === Articles === * [https://www.theinformation.com/articles/venture-fomo-hits-robotics-as-young-startup-gets-1-5-billion-valuation Venture FOMO Hits Robotics as Young Startup Gets $1.5 Billion Valuation] 71894f4fc42406204190ec634243c20c700491be 44 31 2024-04-24T02:03:22Z Ben 2 wikitext text/x-wiki Skild is a stealth foundation model startup started by two faculty members from Carnegie Mellon University. === Articles === * [https://www.theinformation.com/articles/venture-fomo-hits-robotics-as-young-startup-gets-1-5-billion-valuation Venture FOMO Hits Robotics as Young Startup Gets $1.5 Billion Valuation] [[Category:Companies]] 02b55281177a97c01027d3a3f0b832d1fa403c4d 0 12 32 2024-04-24T01:58:44Z Ben 2 Created page with "[https://physicalintelligence.company/ Physical Intelligence] is a company based in the Bay Area which is building foundation models for embodied AI." wikitext text/x-wiki [https://physicalintelligence.company/ Physical Intelligence] is a company based in the Bay Area which is building foundation models for embodied AI. 65657d80e72433bdfc4f25656dfe24154690824a 43 32 2024-04-24T02:02:54Z Ben 2 wikitext text/x-wiki [https://physicalintelligence.company/ Physical Intelligence] is a company based in the Bay Area which is building foundation models for embodied AI. [[Category:Companies]] cb8cec216f584d2700508bd0e832b7c064b4654a 0 13 34 2024-04-24T02:00:11Z Ben 2 Created page with "This is a course taught by Russ Tedrake at MIT." wikitext text/x-wiki This is a course taught by Russ Tedrake at MIT. 1a007d96944837db3fbfa2a82f8c34680a013ed6 14 14 42 2024-04-24T02:02:31Z Ben 2 Created page with "This category is for companies building humanoid robots." wikitext text/x-wiki This category is for companies building humanoid robots. 6074c836c244d3c7c2860678d324d2f31522adb9 0 15 47 2024-04-24T02:04:01Z Ben 2 Ben moved page [[K-Scale Labs Hardware]] to [[Stompy]] wikitext text/x-wiki #REDIRECT [[Stompy]] edf35715d7c99162d2c3854e57ff75dd4cf77d72 0 16 51 2024-04-24T02:30:13Z Ben 2 Created page with "The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. To get onboarded, you should send us the public key..." wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly: <syntaxhighlight> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/is_rsa </syntaxhighlight> 945034cd4748e3fd2e078b3f1d6fdb50e24e6625 52 51 2024-04-24T02:33:08Z Ben 2 wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly: <syntaxhighlight> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/is_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the `/ephemeral` directory. Your home directory should come with a symlink to a subdirectory which you have write access to. f7db11976c42bca986f14192ecbc4a01d37e11c5 53 52 2024-04-24T02:33:24Z Ben 2 wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly: <syntaxhighlight> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/is_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the `/ephemeral` directory. Your home directory should come with a symlink to a subdirectory which you have write access to. 427efcb0ef3deccc8e29da0273398c324c2aa4d3 54 53 2024-04-24T02:33:53Z Ben 2 wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/is_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the `/ephemeral` directory. Your home directory should come with a symlink to a subdirectory which you have write access to. e4bd67c8320b0b135a64925a336efd704144f509 4 17 55 2024-04-24T02:52:32Z Ben 2 Created page with "The Humanoid Robots Wiki is a free public wiki containing general information about humanoid robots. It is maintained by [https://kscale.dev/ K-Scale Labs]." wikitext text/x-wiki The Humanoid Robots Wiki is a free public wiki containing general information about humanoid robots. It is maintained by [https://kscale.dev/ K-Scale Labs]. ce6f28ed550a747d87c969bc3a10eddf85809b47 0 18 56 2024-04-24T02:58:52Z Ben 2 Created page with "Isaac Sim is a simulator from Nvidia based on Omniverse. === Doing Simple Operations === '''Start Isaac Sim''' * Open Omniverse Launcher * Navigate to the Library * Under..." wikitext text/x-wiki Isaac Sim is a simulator from Nvidia based on Omniverse. === Doing Simple Operations === '''Start Isaac Sim''' * Open Omniverse Launcher * Navigate to the Library * Under “Apps” click “Isaac Sim” * Click “Launch” * There are multiple options for launching. Choose the normal one to show the GUI or headless if streaming. * Choose <code>File > Open...</code> and select the <code>.usd</code> model corresponding to the robot you want to simulate. '''Connecting streaming client''' * Start Isaac Sim in Headless (Native) mode * Open Omniverse Streaming Client * Connect to the server [[Category: Simulators]] 551ba3322c3eabc314b546f01ea69381ffd7e22a 14 19 57 2024-04-24T02:59:12Z Ben 2 Created page with "Category for various simulators that you can use" wikitext text/x-wiki Category for various simulators that you can use bd18a7545f0c9a81b86239006203a510d0ede019 0 20 58 2024-04-24T03:07:54Z Ben 2 Created page with "The Jetson Orin is a development board from Nvidia. === CAN Bus === See [https://docs.nvidia.com/jetson/archives/r34.1/DeveloperGuide/text/HR/ControllerAreaNetworkCan.html h..." wikitext text/x-wiki The Jetson Orin is a development board from Nvidia. === CAN Bus === See [https://docs.nvidia.com/jetson/archives/r34.1/DeveloperGuide/text/HR/ControllerAreaNetworkCan.html here] for notes on configuring the CAN bus for the Jetson. Install dependencies: <syntaxhighlight lang="bash"> sudo add-apt-repository ppa:deadsnakes/ppa sudo apt update sudo apt upgrade sudo apt install g++ python3.11-dev </syntaxhighlight> Initialize the CAN bus on startup: <syntaxhighlight lang="bash"> #!/bin/bash # Set pinmux. busybox devmem 0x0c303000 32 0x0000C400 busybox devmem 0x0c303008 32 0x0000C458 busybox devmem 0x0c303010 32 0x0000C400 busybox devmem 0x0c303018 32 0x0000C458 # Install modules. modprobe can modprobe can_raw modprobe mttcan # Turn off CAN. ip link set down can0 ip link set down can1 # Set parameters. ip link set can0 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off ip link set can1 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off # Turn on CAN. ip link set up can0 ip link set up can1 </syntaxhighlight> You can run this script automatically on startup by writing a service configuration to (for example) <code>/etc/systemd/system/can_setup.service</code> <syntaxhighlight lang="text"> [Unit] Description=Initialize CAN Interfaces After=network.target [Service] Type=oneshot ExecStart=/opt/kscale/enable_can.sh RemainAfterExit=true [Install] WantedBy=multi-user.target </syntaxhighlight> To enable this, run: <syntaxhighlight lang="bash"> sudo systemctl enable can_setup sudo systemctl start can_setup </syntaxhighlight> === Cameras === ==== Arducam IMX 219 ==== * [https://www.arducam.com/product/arducam-imx219-multi-camera-kit-for-the-nvidia-jetson-agx-orin/ Product Page] * Shipping was pretty fast * Order a couple backup cameras because a couple of the cameras that they shipped came busted * [https://docs.arducam.com/Nvidia-Jetson-Camera/Nvidia-Jetson-Orin-Series/NVIDIA-Jetson-AGX-Orin/Quick-Start-Guide/ Quick start guide] Run the installation script: <syntaxhighlight lang="bash"> wget https://github.com/ArduCAM/MIPI_Camera/releases/download/v0.0.3/install_full.sh chmod u+x install_full.sh ./install_full.sh -m imx219 </syntaxhighlight> Supported kernel versions (see releases [https://github.com/ArduCAM/MIPI_Camera/releases here]): * <code>5.10.104-tegra-35.3.1</code> * <code>5.10.120-tegra-35.4.1</code> Install an older kernel from [https://developer.nvidia.com/embedded/jetson-linux-archive here]. This required downgrading to Ubuntu 20.04 (only changing <code>/etc/os-version</code>). Install dependencies: <syntaxhighlight lang="bash"> sudo apt update sudo apt install \ gstreamer1.0-tools \ gstreamer1.0-alsa \ gstreamer1.0-plugins-base \ gstreamer1.0-plugins-good \ gstreamer1.0-plugins-bad \ gstreamer1.0-plugins-ugly \ gstreamer1.0-libav sudo apt install libgstreamer1.0-dev \ libgstreamer-plugins-base1.0-dev \ libgstreamer-plugins-good1.0-dev \ libgstreamer-plugins-bad1.0-dev sudo apt install \ v4l-utils \ ffmpeg </syntaxhighlight> Make sure the camera shows up: <syntaxhighlight lang="bash"> v4l2-ctl --list-formats-ext </syntaxhighlight> Capture a frame from the camera: <syntaxhighlight lang="bash"> gst-launch-1.0 nvarguscamerasrc sensor-id=0 ! "video/x-raw(memory:NVMM), width=1280, height=720, framerate=60/1" ! nvvidconv ! jpegenc snapshot=TRUE ! filesink location=test.jpg </syntaxhighlight> Alternatively, use the following Python code: <syntaxhighlight lang="bash"> import cv2 gst_str = ( 'nvarguscamerasrc sensor-id=0 ! ' 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)720, format=(string)NV12, framerate=(fraction)60/1 ! ' 'nvvidconv flip-method=0 ! ' 'video/x-raw, width=(int)1280, height=(int)720, format=(string)BGRx ! ' 'videoconvert ! ' 'video/x-raw, format=(string)BGR ! ' 'appsink' ) cap = cv2.VideoCapture(gst_str, cv2.CAP_GSTREAMER) while True: ret, frame = cap.read() if ret: print(frame) if cv2.waitKey(1) & 0xFF == ord('q'): break else: break cap.release() cv2.destroyAllWindows() </syntaxhighlight> === IMU === We're using the [https://ozzmaker.com/product/berryimu-accelerometer-gyroscope-magnetometer-barometricaltitude-sensor/ BerryIMU v3]. To use it, connect pin 3 on the Jetson to SDA and pin 5 to SCL for I2C bus 7. You can verify the connection is successful if the following command matches: <syntaxhighlight lang="bash"> $ sudo i2cdetect -y -r 7 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- 1c -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- 6a -- -- -- -- -- 70: -- -- -- -- -- -- -- 77 </syntaxhighlight> The equivalent command on the Raspberry Pi should use bus 1: <syntaxhighlight lang="bash"> sudo i2cdetect -y -r 1 </syntaxhighlight> The default addresses are: * <code>0x6A</code>: Gyroscope and accelerometer * <code>0x1C</code>: Magnetometer * <code>0x77</code>: Barometer 3c2dda225f11070061ff5eefce5fdbf5d68061f9 59 58 2024-04-24T03:10:01Z Ben 2 /* Arducam IMX 219 */ wikitext text/x-wiki The Jetson Orin is a development board from Nvidia. === CAN Bus === See [https://docs.nvidia.com/jetson/archives/r34.1/DeveloperGuide/text/HR/ControllerAreaNetworkCan.html here] for notes on configuring the CAN bus for the Jetson. Install dependencies: <syntaxhighlight lang="bash"> sudo add-apt-repository ppa:deadsnakes/ppa sudo apt update sudo apt upgrade sudo apt install g++ python3.11-dev </syntaxhighlight> Initialize the CAN bus on startup: <syntaxhighlight lang="bash"> #!/bin/bash # Set pinmux. busybox devmem 0x0c303000 32 0x0000C400 busybox devmem 0x0c303008 32 0x0000C458 busybox devmem 0x0c303010 32 0x0000C400 busybox devmem 0x0c303018 32 0x0000C458 # Install modules. modprobe can modprobe can_raw modprobe mttcan # Turn off CAN. ip link set down can0 ip link set down can1 # Set parameters. ip link set can0 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off ip link set can1 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off # Turn on CAN. ip link set up can0 ip link set up can1 </syntaxhighlight> You can run this script automatically on startup by writing a service configuration to (for example) <code>/etc/systemd/system/can_setup.service</code> <syntaxhighlight lang="text"> [Unit] Description=Initialize CAN Interfaces After=network.target [Service] Type=oneshot ExecStart=/opt/kscale/enable_can.sh RemainAfterExit=true [Install] WantedBy=multi-user.target </syntaxhighlight> To enable this, run: <syntaxhighlight lang="bash"> sudo systemctl enable can_setup sudo systemctl start can_setup </syntaxhighlight> === Cameras === ==== Arducam IMX 219 ==== * [https://www.arducam.com/product/arducam-imx219-multi-camera-kit-for-the-nvidia-jetson-agx-orin/ Product Page] ** Shipping was pretty fast ** Order a couple backup cameras because a couple of the cameras that they shipped came busted * [https://docs.arducam.com/Nvidia-Jetson-Camera/Nvidia-Jetson-Orin-Series/NVIDIA-Jetson-AGX-Orin/Quick-Start-Guide/ Quick start guide] Run the installation script: <syntaxhighlight lang="bash"> wget https://github.com/ArduCAM/MIPI_Camera/releases/download/v0.0.3/install_full.sh chmod u+x install_full.sh ./install_full.sh -m imx219 </syntaxhighlight> Supported kernel versions (see releases [https://github.com/ArduCAM/MIPI_Camera/releases here]): * <code>5.10.104-tegra-35.3.1</code> * <code>5.10.120-tegra-35.4.1</code> Install an older kernel from [https://developer.nvidia.com/embedded/jetson-linux-archive here]. This required downgrading to Ubuntu 20.04 (only changing <code>/etc/os-version</code>). Install dependencies: <syntaxhighlight lang="bash"> sudo apt update sudo apt install \ gstreamer1.0-tools \ gstreamer1.0-alsa \ gstreamer1.0-plugins-base \ gstreamer1.0-plugins-good \ gstreamer1.0-plugins-bad \ gstreamer1.0-plugins-ugly \ gstreamer1.0-libav sudo apt install libgstreamer1.0-dev \ libgstreamer-plugins-base1.0-dev \ libgstreamer-plugins-good1.0-dev \ libgstreamer-plugins-bad1.0-dev sudo apt install \ v4l-utils \ ffmpeg </syntaxhighlight> Make sure the camera shows up: <syntaxhighlight lang="bash"> v4l2-ctl --list-formats-ext </syntaxhighlight> Capture a frame from the camera: <syntaxhighlight lang="bash"> gst-launch-1.0 nvarguscamerasrc sensor-id=0 ! "video/x-raw(memory:NVMM), width=1280, height=720, framerate=60/1" ! nvvidconv ! jpegenc snapshot=TRUE ! filesink location=test.jpg </syntaxhighlight> Alternatively, use the following Python code: <syntaxhighlight lang="bash"> import cv2 gst_str = ( 'nvarguscamerasrc sensor-id=0 ! ' 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)720, format=(string)NV12, framerate=(fraction)60/1 ! ' 'nvvidconv flip-method=0 ! ' 'video/x-raw, width=(int)1280, height=(int)720, format=(string)BGRx ! ' 'videoconvert ! ' 'video/x-raw, format=(string)BGR ! ' 'appsink' ) cap = cv2.VideoCapture(gst_str, cv2.CAP_GSTREAMER) while True: ret, frame = cap.read() if ret: print(frame) if cv2.waitKey(1) & 0xFF == ord('q'): break else: break cap.release() cv2.destroyAllWindows() </syntaxhighlight> === IMU === We're using the [https://ozzmaker.com/product/berryimu-accelerometer-gyroscope-magnetometer-barometricaltitude-sensor/ BerryIMU v3]. To use it, connect pin 3 on the Jetson to SDA and pin 5 to SCL for I2C bus 7. You can verify the connection is successful if the following command matches: <syntaxhighlight lang="bash"> $ sudo i2cdetect -y -r 7 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- 1c -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- 6a -- -- -- -- -- 70: -- -- -- -- -- -- -- 77 </syntaxhighlight> The equivalent command on the Raspberry Pi should use bus 1: <syntaxhighlight lang="bash"> sudo i2cdetect -y -r 1 </syntaxhighlight> The default addresses are: * <code>0x6A</code>: Gyroscope and accelerometer * <code>0x1C</code>: Magnetometer * <code>0x77</code>: Barometer 7383dadc24c397a0fc8b98ee3286da5213c2d35c 0 1 60 50 2024-04-24T03:32:41Z 92.107.67.203 0 /* Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === ==== Robots ==== * [[Optimus]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Fourier Intelligence]] * [[AGIBot]] * [[Unitree]] * [[UBTech]] * [[Boston Dynamics]] * [[Apptronik]] * [[K-Scale Labs]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] f70c52c93acd7cefbe73d74a535910ac83a03f48 61 60 2024-04-24T03:33:21Z 67.194.230.174 0 /* Learning */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === ==== Robots ==== * [[Optimus]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Fourier Intelligence]] * [[AGIBot]] * [[Unitree]] * [[UBTech]] * [[Boston Dynamics]] * [[Apptronik]] * [[K-Scale Labs]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] * [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] 618df699325fccee847a02377d9d13f43c52a055 62 61 2024-04-24T03:33:23Z 92.107.67.203 0 /* Companies */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === * [[Tesla]] ==== Robots ==== * [[Optimus]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Fourier Intelligence]] * [[AGIBot]] * [[Unitree]] * [[UBTech]] * [[Boston Dynamics]] * [[Apptronik]] * [[K-Scale Labs]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] * [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] 28f71b8edd069d183f2c657f8502e6df350c3748 66 62 2024-04-24T03:56:08Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === * [[Tesla]] * [[K-Scale Labs]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Fourier Intelligence]] * [[AGIBot]] * [[Unitree]] * [[UBTech]] * [[Boston Dynamics]] * [[Apptronik]] ==== Robots ==== * [[Optimus]] * [[Unitree H1]] * [[Neo]] * [[Eve]] * [[Stompy]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] * [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] 08aa494327cb0a5746ce6aee42ea18954f7d75e6 67 66 2024-04-24T03:56:15Z Ben 2 /* Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Companies === * [[Tesla]] * [[K-Scale Labs]] * [[Agility]] * [[Sanctuary]] * [[1X]] * [[Fourier Intelligence]] * [[AGIBot]] * [[Unitree]] * [[UBTech]] * [[Boston Dynamics]] * [[Apptronik]] ==== Robots ==== * [[Optimus]] * [[H1]] * [[Neo]] * [[Eve]] * [[Stompy]] ==== Foundation Models ==== * [[Physical Intelligence]] * [[Skild]] === Learning === * [[Underactuated Robotics]] * [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] 6f2242e4a681e01cfe54cc392fa93356938ecc00 68 67 2024-04-24T04:00:16Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | |- | [[Sanctuary]] | |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | |- | [[Unitree]] | [[Unitree H1]] |- | [[UBTech]] | |- | [[Boston Dynamics]] | [[Atlas]] |- | [[Apptronik]] | |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} === Learning === * [[Underactuated Robotics]] * [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] 92bbc2f383bb9bc022f6e5b9d976c97e5d80fe8a 69 68 2024-04-24T04:02:35Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | |- | [[Sanctuary]] | |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | |- | [[Unitree]] | [[Unitree H1]] |- | [[UBTech]] | |- | [[Boston Dynamics]] | [[Atlas]] |- | [[Apptronik]] | |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 59f6537df964aca7ba0aa870aaca342885d15abe 70 69 2024-04-24T04:07:22Z 76.144.71.131 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3], [Digit]] | |- | [[Sanctuary]] | |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | |- | [[Unitree]] | [[Unitree H1]] |- | [[UBTech]] | |- | [[Boston Dynamics]] | [[Atlas]] |- | [[Apptronik]] | |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 29aabc00229d86fc9577af5b99de2951c5d6f5a7 71 70 2024-04-24T04:07:39Z 76.144.71.131 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] | |- | [[Sanctuary]] | |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | |- | [[Unitree]] | [[Unitree H1]] |- | [[UBTech]] | |- | [[Boston Dynamics]] | [[Atlas]] |- | [[Apptronik]] | |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 5e544a353e56719d0258e928406ef477dc66998d 72 71 2024-04-24T04:08:03Z 76.144.71.131 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | |- | [[Unitree]] | [[Unitree H1]] |- | [[UBTech]] | |- | [[Boston Dynamics]] | [[Atlas]] |- | [[Apptronik]] | |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 406e775c8d26bee782751732b54e55b70e8264fa 73 72 2024-04-24T04:11:14Z 76.144.71.131 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[Unitree H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 418c6beda54489deea97092a203f01e389757d4f 77 73 2024-04-24T04:16:04Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} bdffb97a65fde89194f2771b67861bb26b8662aa 78 77 2024-04-24T04:19:44Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[FDROBOT]] | [[T1]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 72be79b91b2579f2c44587261a43d18b5f534486 80 78 2024-04-24T04:21:51Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot-L]], [[XBot-S]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 5e0a77ade68bcef5f2b385c697cb306ef699adf2 81 80 2024-04-24T04:22:02Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 6fc7f12cf7bf6efed3741753a06f36fbf48e5270 82 81 2024-04-24T04:26:19Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 01fd23f8b4684298e0e6679be5792cf48137ce0e 83 82 2024-04-24T04:30:38Z 172.56.152.64 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} d25f9d513123395775976810013e0e7865801797 96 83 2024-04-24T04:50:24Z 76.144.71.131 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]],[[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 5c2d700f58628e341ca655b59a338a50973c9d76 97 96 2024-04-24T04:50:43Z 76.144.71.131 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 2008cf132cfd5d97e9ba68f8fe699cd5efd44d35 0 13 63 34 2024-04-24T03:36:12Z Sagar 4 Link to the course website. wikitext text/x-wiki This is a course taught by Russ Tedrake at MIT. https://underactuated.csail.mit.edu/ a5f5b589eedfafe7c56b8fcb50c08bcb9710bfcd 0 21 64 2024-04-24T03:54:46Z Ben 2 Created page with "AR glasses for data capture from Meta. === References === ==== Links ==== * [https://www.projectaria.com/ Website] * [https://docs.ego-exo4d-data.org/ Ego4D Documentation]..." wikitext text/x-wiki AR glasses for data capture from Meta. === References === ==== Links ==== * [https://www.projectaria.com/ Website] * [https://docs.ego-exo4d-data.org/ Ego4D Documentation] * [https://facebookresearch.github.io/projectaria_tools/docs/intro Project Aria Documentation] ** [https://facebookresearch.github.io/projectaria_tools/docs/data_formats/mps/mps_trajectory Specific page regarding trajectories] ==== Datasets ==== * [https://www.projectaria.com/datasets/apd/ APD Dataset] 811d22fc32ee1c2e3a74958f9b447a7b83c40ef8 0 22 65 2024-04-24T03:55:18Z Ben 2 Created page with "The humanoid robot from Tesla. [[Category:Robots]]" wikitext text/x-wiki The humanoid robot from Tesla. [[Category:Robots]] 0887b16263546ec0d33e5f7497481233a726ddc1 0 10 74 38 2024-04-24T04:15:29Z 76.144.71.131 0 wikitext text/x-wiki [https://www.1x.tech/ 1X] (formerly known as Halodi Robotics) is a humanoid robotics company based in Moss, Norway. They have two robots: EVE and NEO. EVE is a wheeled robot while NEO has legs. The company is known for it's high torque BLDC motors that they developed in house. Those BLDC motors are paired with low gear ratio cable drives. EVE and NEO are designed for safe human interaction by reducing actuator inertia. [[Category:Companies]] 6bd3830bc47b96572ede840ac193cc7a832c6611 0 3 75 15 2024-04-24T04:15:52Z Ben 2 Ben moved page [[Unitree H1]] to [[H1]] wikitext text/x-wiki It is available for purchase [https://shop.unitree.com/products/unitree-h1 here]. 900095efde90381d578981a4184f54e668da07cf 0 23 76 2024-04-24T04:15:52Z Ben 2 Ben moved page [[Unitree H1]] to [[H1]] wikitext text/x-wiki #REDIRECT [[H1]] 1abd28e18815dc95839514830b8b3c4cc11ceef6 0 24 79 2024-04-24T04:21:43Z 76.144.71.131 0 Created page with " Cassie is a bipedal robot designed by Oregon State University and licensed and built by Agility Robotics. ==Development==" wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by Agility Robotics. ==Development== 95fef9ac2606c58a82094bf7a30d09311af26bdd 84 79 2024-04-24T04:31:28Z 76.144.71.131 0 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by Agility Robotics. ==Specifications== {| class="wikitable" |- | Height || 115 cm |- | Weight|| 31 kg |- | Speed || >4 m/s (8.95 mph) |- | Payload || Example |- | Battery Life || ~5 hours |- | Battery Capacity || 1 kWh |- | DoF || 10 (5 per leg) |- | Cost || $250,000 |- | Number Made || ~12 |- | Status || Retired |} ==Development== ==World Record== 41764868683af6d4cc1bc0dbf92702c13203a12d 89 84 2024-04-24T04:44:38Z 76.144.71.131 0 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by Agility Robotics. ==Specifications== {| class="wikitable" |- | Height || 115 cm |- | Weight|| 31 kg |- | Speed || >4 m/s (8.95 mph) |- | Payload || Example |- | Battery Life || ~5 hours |- | Battery Capacity || 1 kWh |- | DoF || 10 (5 per leg) |- | Cost || $250,000 |- | Number Made || ~12 |- | Status || Retired |} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. 77c6d661dbaff15a99c517f5d8b5cc61cc742dd5 10 25 85 2024-04-24T04:36:50Z Ben 2 Created page with "<includeonly>{{Infobox | titlestyle = padding-bottom:0.25em | title = {{If empty |{{{name|}}} |{{{shorttitle|}}} |{{PAGENAMEBASE}} }} | headerstyle = background:#bbddff | la..." wikitext text/x-wiki <includeonly>{{Infobox | titlestyle = padding-bottom:0.25em | title = {{If empty |{{{name|}}} |{{{shorttitle|}}} |{{PAGENAMEBASE}} }} | headerstyle = background:#bbddff | labelstyle = padding-top:0.245em;line-height:1.15em;padding-right:0.65em | datastyle = text-align:left;line-height:1.3em | autoheaders = y | label1 = Legal Name | data1 = {{{legal_name|}}} | label2 = Country | data2 = {{{country|}}} }}</includeonly> 20553278eaca4b3ec4b13f68444fc675357d21f7 91 85 2024-04-24T04:46:56Z Admin 1 wikitext text/x-wiki <includeonly> <table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name|Infobox}}}</th></tr> <tr>{{{content}}}</tr> </table> </includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> b3f9201ace7e3b9f58308ad3ad7153272d574e0c 92 91 2024-04-24T04:48:13Z Admin 1 wikitext text/x-wiki <includeonly> <table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name|Infobox}}}</th></tr> <tr><td>{{{legal_name}}}</td></tr> <tr><td>{{{robots}}}</td></tr> </table> </includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> 3cc7fdfc0c7d3cfe77f1e80b76152d4f9b5edf87 94 92 2024-04-24T04:49:14Z Admin 1 wikitext text/x-wiki <includeonly> <table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name|Infobox}}}</th></tr> <tr><td>{{{legal_name}}}</td></tr> <tr><td>{{{country}}}</td></tr> <tr><td>{{{robots}}}</td></tr> </table> </includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> 0f52a825209243299bf1c124c714997a35041f9e 98 94 2024-04-24T04:51:28Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name|Infobox}}}</th></tr> <tr><td>{{{legal_name}}}</td></tr> <tr><td>{{{country}}}</td></tr> <tr><td>{{{robots}}}</td></tr> </table></includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> 7417fe1483e2208ad946d92b963d263a2f2b6148 0 7 86 41 2024-04-24T04:37:02Z Ben 2 wikitext text/x-wiki {{infobox company | legal_name = Tesla, Inc. | country = United States }} Tesla is building a humanoid robot called Optimus. [[Category:Companies]] cdd4bd231071a526b5864e5095c2955b83dc917a 90 86 2024-04-24T04:45:57Z Admin 1 wikitext text/x-wiki {{infobox | name = Tesla, Inc. | content = United States }} Tesla is building a humanoid robot called Optimus. [[Category:Companies]] 378f21a0d043b7d9bc2f2548e3ed395a91799e4f 93 90 2024-04-24T04:48:57Z Admin 1 wikitext text/x-wiki {{infobox | name = Tesla | legal_name = Tesla, Inc. | country = United States }} Tesla is building a humanoid robot called Optimus. [[Category:Companies]] 568833acd44f8fdbf14c9cfe6d375db41881e0c4 95 93 2024-04-24T04:49:27Z Admin 1 wikitext text/x-wiki {{infobox company | name = Tesla | legal_name = Tesla, Inc. | country = United States | robots = [[Optimus]] }} Tesla is building a humanoid robot called Optimus. [[Category:Companies]] 4000c6e462c12c25bed24bd2773822570704c918 99 95 2024-04-24T04:51:39Z Admin 1 wikitext text/x-wiki {{infobox company | name = Tesla | legal_name = Tesla, Inc. | country = United States | robots = [[Optimus]] }} Tesla is building a humanoid robot [[Optimus]]. [[Category:Companies]] d7a1a1c9a0cb01ab77d5a3c95645b62a49c9afc2 8 27 88 2024-04-24T04:43:29Z Admin 1 Created page with "/* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; /* or any othe..." css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; /* or any other width */ } .infobox th { background-color: #e0e0e0; text-align: center; } .infobox td { padding: 2px 5px; } caa9d3efdd5b6e6240c159cca1491176282aa468 10 28 100 2024-04-24T04:54:35Z Admin 1 Created page with "<includeonly><table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr> <tr>..." wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr> <tr><td>{{{company}}}</td></tr> <tr><td>{{{height}}}</td></tr> <tr><td>{{{weight}}}</td></tr> <tr><td>{{{single_hand_payload}}}</td></tr> <tr><td>{{{two_hand_payload}}}</td></tr> <tr><td>{{{cost}}}</td></tr> </table></includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> 0bcfbc3b04b0614bbf224da73c057cab969de81f 101 100 2024-04-24T04:55:10Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr> <tr><td>{{{organization}}}</td></tr> <tr><td>{{{height}}}</td></tr> <tr><td>{{{weight}}}</td></tr> <tr><td>{{{single_hand_payload}}}</td></tr> <tr><td>{{{two_hand_payload}}}</td></tr> <tr><td>{{{cost}}}</td></tr> </table></includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> e13df0ffdf3cc2f2ca183847408c06d3ce3ff6e5 0 22 102 65 2024-04-24T04:55:27Z Admin 1 wikitext text/x-wiki {{infobox robot | name = Optimus | organization = [[Tesla]] }} The humanoid robot from Tesla. [[Category:Robots]] 07c310e4aa39cda84ad4ce3cc810fc0c24a7cd63 103 102 2024-04-24T04:58:51Z Admin 1 wikitext text/x-wiki {{infobox robot | name = Optimus | organization = [[Tesla]] }} The humanoid robot from [[Tesla]]. [[Category:Robots]] 49bda9acc845a832bea425b1381eb3eeb3cc279b 108 103 2024-04-24T05:14:00Z Admin 1 wikitext text/x-wiki {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} The humanoid robot from [[Tesla]]. [[Category:Robots]] aee31315ff860dfa597ff905681816578e199bb5 110 108 2024-04-24T05:16:54Z Admin 1 wikitext text/x-wiki {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} Optimus is a humanoid robot from [[Tesla]]. [[Category:Robots]] 653e328581e8a58e1692ee34f1bb0bd3fc120038 111 110 2024-04-24T05:17:43Z Admin 1 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} [[Category:Robots]] 18d8db29eee0ad16cd78cc79ae7e2ddbd64224a4 113 111 2024-04-24T05:18:49Z Admin 1 wikitext text/x-wiki {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} Optimus is a humanoid robot from [[Tesla]]. [[Category:Robots]] 653e328581e8a58e1692ee34f1bb0bd3fc120038 119 113 2024-04-24T05:23:46Z Admin 1 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} [[Category:Robots]] 18d8db29eee0ad16cd78cc79ae7e2ddbd64224a4 125 119 2024-04-24T06:32:59Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} [[File:Optimus Tesla.jpg|thumb|The Optimus robot from Tesla]] [[Category:Robots]] ef4acfeea1a71412d92323c84f3db70ca706f2c2 126 125 2024-04-24T06:34:44Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} [[Category:Robots]] 18d8db29eee0ad16cd78cc79ae7e2ddbd64224a4 129 126 2024-04-24T06:37:23Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} [[File:Optimus Tesla (1).jpg|thumb]] [[Category:Robots]] 08153e0cb7951922620fdef5b32073bfd293dc31 133 129 2024-04-24T06:55:29Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 01c380222326d68e1ceb000aab6e6484211ed5b1 135 133 2024-04-24T07:02:22Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc | purchase_link = Rumored }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] a66c1514cc28f472cbaf2b1cf58ac4348b9ed72c 137 135 2024-04-24T07:03:36Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc | purchase_link = Rumored 2025 }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 29a8b1ffb43136e8ce13e9e6bb533b551e2146a3 140 137 2024-04-24T07:04:18Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc | purchase_link = Rumored, 2025 }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 2728ad91f8489a5010db108d676fa2eeb0301548 144 140 2024-04-24T07:06:20Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc | purchase_link = Rumored 2025 }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 29a8b1ffb43136e8ce13e9e6bb533b551e2146a3 150 144 2024-04-24T07:14:37Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 01c380222326d68e1ceb000aab6e6484211ed5b1 14 29 104 2024-04-24T04:59:26Z Admin 1 Created page with "Category for specific humanoid robot implementations." wikitext text/x-wiki Category for specific humanoid robot implementations. f2dd3ff9b9d3216304e77dd99677a66b91b3a8bc 14 30 105 2024-04-24T04:59:53Z Admin 1 Created page with "Category for templates that are used in various places." wikitext text/x-wiki Category for templates that are used in various places. 47f74cd9e844eb53ef5683662fd622a553b56ba6 10 28 106 101 2024-04-24T05:12:02Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr> {{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }} {{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }} {{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }} {{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }} {{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }} {{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }} </table></includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> cb414c4239fecb4c16e44d0d08b7c0c4c11c6a45 107 106 2024-04-24T05:13:54Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr> {{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }} {{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }} {{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }} {{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }} {{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }} {{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }} {{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }} </table></includeonly> <noinclude> This is the template for a basic infobox. [[Category:Templates]] </noinclude> 1a9eda5c3d90bf0bbab5563696c5f6de4cd93875 112 107 2024-04-24T05:18:34Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"><tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr>{{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }}{{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }}{{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }}{{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }}{{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }}{{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }}{{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }}</table></includeonly><noinclude>This is the template for a basic infobox. [[Category:Templates]]</noinclude> 856023026e283d367074329b9a713883adc140b3 114 112 2024-04-24T05:20:20Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"><tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr>{{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }}{{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }}{{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }}{{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }}{{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }}{{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }}{{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }}</table></includeonly><noinclude>This is the template for a basic infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> [[Category:Templates]]</noinclude> 10bb3d1356be6b0fec5bff320448ec3d877b9268 116 114 2024-04-24T05:22:34Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"><tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr>{{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }}{{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }}{{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }}{{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }}{{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }}{{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }}{{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }}</table></includeonly><noinclude>This is the template for the robot infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> [[Category:Templates]]</noinclude> 66c5e2ab61fc431d8afb87d57343038f813ca717 134 116 2024-04-24T07:01:52Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"><tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr>{{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }}{{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }}{{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }}{{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }}{{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }}{{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }}{{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }}{{#if: {{{purchase_link|}}} | <tr><th>Purchase Link</th><td>{{{purchase_link}}}</td></tr> }}</table></includeonly><noinclude>This is the template for the robot infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> * <code>purchase_link</code> [[Category:Templates]]</noinclude> b5a0532191b1ca99b8c06d91e9745d183048c1fd 136 134 2024-04-24T07:03:25Z Mrroboto 5 wikitext text/x-wiki <includeonly> <table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr {{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }} {{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }} {{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }} {{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }} {{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }} {{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }} {{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }} {{#if: {{{purchase_link|}}} | <tr><th>Purchase Link</th><td>{{{purchase_link}}}</td></tr> }} </table> </includeonly><noinclude>This is the template for the robot infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> * <code>purchase_link</code> [[Category:Templates]]</noinclude> 2714fd9559872125052b747da6323e188dc464b6 138 136 2024-04-24T07:03:49Z Mrroboto 5 wikitext text/x-wiki <includeonly> <table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr> {{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }} {{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }} {{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }} {{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }} {{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }} {{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }} {{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }} {{#if: {{{purchase_link|}}} | <tr><th>Purchase Link</th><td>{{{purchase_link}}}</td></tr> }} </table> </includeonly><noinclude>This is the template for the robot infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> * <code>purchase_link</code> [[Category:Templates]]</noinclude> 1118d0ed157ba26842b4942040cfdb1d9ef5ae8e 139 138 2024-04-24T07:04:10Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"> <tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr> {{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }} {{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }} {{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }} {{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }} {{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }} {{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }} {{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }} {{#if: {{{purchase_link|}}} | <tr><th>Purchase Link</th><td>{{{purchase_link}}}</td></tr> }} </table></includeonly><noinclude>This is the template for the robot infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> * <code>purchase_link</code> [[Category:Templates]]</noinclude> 0e3d0588acc6bd9392db7d8bd73df0ad8843549f 143 139 2024-04-24T07:06:14Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox"> <tr><th colspan="2">{{{name}}}</th></tr> {{#if: {{{organization|}}} | <tr><th>Organization</th><td>{{{organization}}}</td></tr> }} {{#if: {{{height|}}} | <tr><th>Height</th><td>{{{height}}}</td></tr> }} {{#if: {{{weight|}}} | <tr><th>Weight</th><td>{{{weight}}}</td></tr> }} {{#if: {{{single_hand_payload|}}} | <tr><th>Single Hand Payload</th><td>{{{single_hand_payload}}}</td></tr> }} {{#if: {{{two_hand_payload|}}} | <tr><th>Two Hand Payload</th><td>{{{two_hand_payload}}}</td></tr> }} {{#if: {{{cost|}}} | <tr><th>Cost</th><td>{{{cost}}}</td></tr> }} {{#if: {{{video|}}} | <tr><th>Video</th><td>{{{video}}}</td></tr> }} {{#if: {{{purchase_link|}}} | <tr><th>Purchase Link</th><td>{{{purchase_link}}}</td></tr> }} </table></includeonly><noinclude>This is the template for the robot infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> * <code>purchase_link</code> [[Category:Templates]]</noinclude> f9e98c710373655935f9b24db331d5cb0839e633 147 143 2024-04-24T07:12:16Z Mrroboto 5 wikitext text/x-wiki {{infobox | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization}}} | key3 = Video | value3 = {{{video}}} | key4 = Purchase Link | value4 = {{{purchase_link}}} }} ba2714a86aa37393cc9b0346c27b6f22eda993e8 148 147 2024-04-24T07:13:06Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization}}} | key3 = Video | value3 = {{{video}}} | key4 = Purchase Link | value4 = {{{purchase_link}}} }} 49435000619f6d642d099c22d85c698d8724f775 149 148 2024-04-24T07:14:18Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization}}} | key3 = Video | value3 = {{#if: {{{video|}}} | [{{{video}}} Video] }} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} }} 6673131dca65c16a87cc3f403e2602c87195dd77 151 149 2024-04-24T07:15:40Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization}}} | key3 = {{#if: {{{video|}}} | Video }} | value3 = [{{{video}}} Video] | key4 = {{#if: {{{purchase_link|}}} | Purchase }} | value4 = [{{{purchase_link}}} Link] }} c5fec06d3992574d88eeb0ed047aa5e87d1ac3f4 8 27 109 88 2024-04-24T05:16:21Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; float: right; margin: 10px 0 10px 10px; } .infobox th { background-color: #e0e0e0; text-align: center; padding: 2px 5px; } .infobox td { padding: 2px 5px; } 8fcebcc4a2b165c433ff7bdc6ecf620967085ded 118 109 2024-04-24T05:23:32Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; } .infobox th { background-color: #e0e0e0; text-align: center; padding: 2px 5px; } .infobox td { padding: 2px 5px; } f49b4037e90fb4b453cf022dfb913fe3f36f34be 120 118 2024-04-24T05:24:05Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; margin-top: 20px; margin-bottom: 20px; } .infobox th { background-color: #e0e0e0; text-align: center; padding: 2px 5px; } .infobox td { padding: 2px 5px; } a904b3fdeefdb2ef478c3ef8b6e850dd88683927 141 120 2024-04-24T07:05:48Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox th { background-color: #e0e0e0; text-align: center; padding: 2px 5px; } .infobox td { padding: 2px 5px; } c965a4d4f15316e124ff6af9fae0e1b7471fe083 142 141 2024-04-24T07:06:08Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox th { background-color: #e0e0e0; text-align: center; padding: 2px 5px; text-align: center; background-color: #cccccc; } .infobox td { padding: 2px 5px; } 546e7039b9abd8427f3adf807b3081cf65c596f4 10 25 115 98 2024-04-24T05:22:25Z Admin 1 wikitext text/x-wiki <includeonly><table class="infobox" style="width: 300px; font-size: 90%;"><tr><th colspan="2" style="text-align: center; background-color: #cccccc;">{{{name}}}</th></tr>{{#if: {{{country|}}} | <tr><th>Country</th><td>{{{country}}}</td></tr> }}{{#if: {{{website|}}} | <tr><th>Website</th><td>{{{website}}}</td></tr> }}{{#if: {{{robots|}}} | <tr><th>Robots</th><td>{{{robots}}}</td></tr> }}</table></includeonly><noinclude>This is the template for the company infobox. Fields are: * <code>name</code> * <code>country</code> * <code>website</code> * <code>robots</code> [[Category:Templates]]</noinclude> 8dcbacb12f7bf1c36d7f01226a71fac02ce0d1a7 0 7 117 99 2024-04-24T05:23:00Z Admin 1 wikitext text/x-wiki {{infobox company | name = Tesla | country = United States | website = https://www.tesla.com/ | robots = [[Optimus]] }} Tesla is building a humanoid robot [[Optimus]]. [[Category:Companies]] c50774886424a0db025582065610a1dcc1b1587d 121 117 2024-04-24T05:24:17Z Admin 1 wikitext text/x-wiki Tesla is building a humanoid robot [[Optimus]]. {{infobox company | name = Tesla | country = United States | website = https://www.tesla.com/ | robots = [[Optimus]] }} [[Category:Companies]] c6fa48baea02ce0575a318a3485e2db5e3a159ad 122 121 2024-04-24T05:24:27Z Admin 1 wikitext text/x-wiki Tesla is building a humanoid robot called [[Optimus]]. {{infobox company | name = Tesla | country = United States | website = https://www.tesla.com/ | robots = [[Optimus]] }} [[Category:Companies]] 9379163b9886449543635e50f77ca3d807f73dc5 0 1 123 97 2024-04-24T06:26:40Z 69.181.66.238 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} c7fc6a3b6e403e5c88a273cecdedff5ae1366758 0 32 127 2024-04-24T06:36:45Z 69.181.66.238 0 Created page with "= Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches..." wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. == Training frameworks == Popular training frameworks ===Isaac Gym=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===Gymnasium=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. == Training algorithms == ===Imitation learning=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===Reinforcement Learning=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. b78ad4ec7eb2be38016e656a01cf1c1a172e25e6 130 127 2024-04-24T06:41:45Z 69.181.66.238 0 /* Reinforcement Learning */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. == Training frameworks == Popular training frameworks ===Isaac Gym=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===Gymnasium=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. == Training algorithms == ===Imitation learning=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. 59b94a0456b67cf00600b3b2af314df3fb5e7183 132 130 2024-04-24T06:44:15Z 69.181.66.238 0 wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks ===Isaac Gym=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===Gymnasium=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. == Training algorithms == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. 22e708af83e215987cfcd7921afea72992718b3b 6 33 128 2024-04-24T06:37:17Z Mrroboto 5 wikitext text/x-wiki The Optimus robot from Tesla 9baab017b6f3bd9f54f2b4d7fe33d8f428602773 0 34 131 2024-04-24T06:42:26Z 69.181.66.238 0 Created page with " ==Training algorithms== ===A2C=== ===PPO===" wikitext text/x-wiki ==Training algorithms== ===A2C=== ===PPO=== 95d6e5ffc9d381d85d89b934739af7387fec932f 10 35 145 2024-04-24T07:09:11Z Mrroboto 5 Created page with "<includeonly><table class="infobox"><tr><th colspan="2">{{{name}}}</th></tr>{{#if: {{{key1|}}} | <tr><th>{{{key1}}}</th><td>{{{value1}}}</td></tr> }}{{#if: {{{key2|}}} | <tr><..." wikitext text/x-wiki <includeonly><table class="infobox"><tr><th colspan="2">{{{name}}}</th></tr>{{#if: {{{key1|}}} | <tr><th>{{{key1}}}</th><td>{{{value1}}}</td></tr> }}{{#if: {{{key2|}}} | <tr><th>{{{key2}}}</th><td>{{{value2}}}</td></tr> }}{{#if: {{{key3|}}} | <tr><th>{{{key3}}}</th><td>{{{value3}}}</td></tr> }}{{#if: {{{key4|}}} | <tr><th>{{{key4}}}</th><td>{{{value4}}}</td></tr> }}{{#if: {{{key5|}}} | <tr><th>{{{key5}}}</th><td>{{{value5}}}</td></tr> }}{{#if: {{{key6|}}} | <tr><th>{{{key6}}}</th><td>{{{value6}}}</td></tr> }}{{#if: {{{key7|}}} | <tr><th>{{{key7}}}</th><td>{{{value7}}}</td></tr> }}{{#if: {{{key8|}}} | <tr><th>{{{key8}}}</th><td>{{{value8}}}</td></tr> }}{{#if: {{{key9|}}} | <tr><th>{{{key9}}}</th><td>{{{value9}}}</td></tr> }}{{#if: {{{key10|}}} | <tr><th>{{{key10}}}</th><td>{{{value10}}}</td></tr> }}{{#if: {{{key11|}}} | <tr><th>{{{key11}}}</th><td>{{{value11}}}</td></tr> }}{{#if: {{{key12|}}} | <tr><th>{{{key12}}}</th><td>{{{value12}}}</td></tr> }}{{#if: {{{key13|}}} | <tr><th>{{{key13}}}</th><td>{{{value13}}}</td></tr> }}{{#if: {{{key14|}}} | <tr><th>{{{key14}}}</th><td>{{{value14}}}</td></tr> }}{{#if: {{{key15|}}} | <tr><th>{{{key15}}}</th><td>{{{value15}}}</td></tr> }}{{#if: {{{key16|}}} | <tr><th>{{{key16}}}</th><td>{{{value16}}}</td></tr> }}{{#if: {{{key17|}}} | <tr><th>{{{key17}}}</th><td>{{{value17}}}</td></tr> }}{{#if: {{{key18|}}} | <tr><th>{{{key18}}}</th><td>{{{value18}}}</td></tr> }}{{#if: {{{key19|}}} | <tr><th>{{{key19}}}</th><td>{{{value19}}}</td></tr> }}{{#if: {{{key20|}}} | <tr><th>{{{key20}}}</th><td>{{{value20}}}</td></tr> }}</table></includeonly><noinclude>This is the template for the robot infobox. Fields are: * <code>name</code> * <code>organization</code> * <code>height</code> * <code>weight</code> * <code>single_hand_payload</code> * <code>two_hand_payload</code> * <code>cost</code> * <code>video</code> * <code>purchase_link</code> [[Category:Templates]]</noinclude> deb580d5370d8c6d4555b8f5c3424a4917d34e97 146 145 2024-04-24T07:10:27Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox"><tr><th colspan="2">{{{name}}}</th></tr>{{#if: {{{key1|}}} | <tr><th>{{{key1}}}</th><td>{{{value1}}}</td></tr> }}{{#if: {{{key2|}}} | <tr><th>{{{key2}}}</th><td>{{{value2}}}</td></tr> }}{{#if: {{{key3|}}} | <tr><th>{{{key3}}}</th><td>{{{value3}}}</td></tr> }}{{#if: {{{key4|}}} | <tr><th>{{{key4}}}</th><td>{{{value4}}}</td></tr> }}{{#if: {{{key5|}}} | <tr><th>{{{key5}}}</th><td>{{{value5}}}</td></tr> }}{{#if: {{{key6|}}} | <tr><th>{{{key6}}}</th><td>{{{value6}}}</td></tr> }}{{#if: {{{key7|}}} | <tr><th>{{{key7}}}</th><td>{{{value7}}}</td></tr> }}{{#if: {{{key8|}}} | <tr><th>{{{key8}}}</th><td>{{{value8}}}</td></tr> }}{{#if: {{{key9|}}} | <tr><th>{{{key9}}}</th><td>{{{value9}}}</td></tr> }}{{#if: {{{key10|}}} | <tr><th>{{{key10}}}</th><td>{{{value10}}}</td></tr> }}{{#if: {{{key11|}}} | <tr><th>{{{key11}}}</th><td>{{{value11}}}</td></tr> }}{{#if: {{{key12|}}} | <tr><th>{{{key12}}}</th><td>{{{value12}}}</td></tr> }}{{#if: {{{key13|}}} | <tr><th>{{{key13}}}</th><td>{{{value13}}}</td></tr> }}{{#if: {{{key14|}}} | <tr><th>{{{key14}}}</th><td>{{{value14}}}</td></tr> }}{{#if: {{{key15|}}} | <tr><th>{{{key15}}}</th><td>{{{value15}}}</td></tr> }}{{#if: {{{key16|}}} | <tr><th>{{{key16}}}</th><td>{{{value16}}}</td></tr> }}{{#if: {{{key17|}}} | <tr><th>{{{key17}}}</th><td>{{{value17}}}</td></tr> }}{{#if: {{{key18|}}} | <tr><th>{{{key18}}}</th><td>{{{value18}}}</td></tr> }}{{#if: {{{key19|}}} | <tr><th>{{{key19}}}</th><td>{{{value19}}}</td></tr> }}{{#if: {{{key20|}}} | <tr><th>{{{key20}}}</th><td>{{{value20}}}</td></tr> }}</table></includeonly><noinclude>This is the template for the robot infobox. Fields are: {{#if: {{{key1|}}} | * <code>{{{key1}}}</code> <code>{{{value1}}}</code> }} {{#if: {{{key2|}}} | * <code>{{{key2}}}</code> <code>{{{value2}}}</code> }} {{#if: {{{key3|}}} | * <code>{{{key3}}}</code> <code>{{{value3}}}</code> }} {{#if: {{{key4|}}} | * <code>{{{key4}}}</code> <code>{{{value4}}}</code> }} {{#if: {{{key5|}}} | * <code>{{{key5}}}</code> <code>{{{value5}}}</code> }} [[Category:Templates]]</noinclude> 69d696dd41b099fd45ff988b79456a23d88e0c7f 152 146 2024-04-24T07:18:21Z Mrroboto 5 wikitext text/x-wiki <includeonly> <table class="infobox"> <tr> <th colspan="2">{{{name}}}</th> </tr>{{#if: {{{key1|}}} | <tr> <td class="infobox-left">{{{key1}}}</td> <td class="infobox-right">{{{value1}}}</td> </tr> }}{{#if: {{{key2|}}} | <tr> <td class="infobox-left">{{{key2}}}</td> <td class="infobox-right">{{{value2}}}</td> </tr> }}{{#if: {{{key3|}}} | <tr> <td class="infobox-left">{{{key3}}}</td> <td class="infobox-right">{{{value3}}}</td> </tr> }}{{#if: {{{key4|}}} | <tr> <td class="infobox-left">{{{key4}}}</td> <td class="infobox-right">{{{value4}}}</td> </tr> }}{{#if: {{{key5|}}} | <tr> <td class="infobox-left">{{{key5}}}</td> <td class="infobox-right">{{{value5}}}</td> </tr> }}{{#if: {{{key6|}}} | <tr> <td class="infobox-left">{{{key6}}}</td> <td class="infobox-right">{{{value6}}}</td> </tr> }}{{#if: {{{key7|}}} | <tr> <td class="infobox-left">{{{key7}}}</td> <td class="infobox-right">{{{value7}}}</td> </tr> }}{{#if: {{{key8|}}} | <tr> <td class="infobox-left">{{{key8}}}</td> <td class="infobox-right">{{{value8}}}</td> </tr> }}{{#if: {{{key9|}}} | <tr> <td class="infobox-left">{{{key9}}}</td> <td class="infobox-right">{{{value9}}}</td> </tr> }}{{#if: {{{key10|}}} | <tr> <td class="infobox-left">{{{key10}}}</td> <td class="infobox-right">{{{value10}}}</td> </tr> }}{{#if: {{{key11|}}} | <tr> <td class="infobox-left">{{{key11}}}</td> <td class="infobox-right">{{{value11}}}</td> </tr> }}{{#if: {{{key12|}}} | <tr> <td class="infobox-left">{{{key12}}}</td> <td class="infobox-right">{{{value12}}}</td> </tr> }}{{#if: {{{key13|}}} | <tr> <td class="infobox-left">{{{key13}}}</td> <td class="infobox-right">{{{value13}}}</td> </tr> }}{{#if: {{{key14|}}} | <tr> <td class="infobox-left">{{{key14}}}</td> <td class="infobox-right">{{{value14}}}</td> </tr> }}{{#if: {{{key15|}}} | <tr> <td class="infobox-left">{{{key15}}}</td> <td class="infobox-right">{{{value15}}}</td> </tr> }}{{#if: {{{key16|}}} | <tr> <td class="infobox-left">{{{key16}}}</td> <td class="infobox-right">{{{value16}}}</td> </tr> }}{{#if: {{{key17|}}} | <tr> <td class="infobox-left">{{{key17}}}</td> <td class="infobox-right">{{{value17}}}</td> </tr> }}{{#if: {{{key18|}}} | <tr> <td class="infobox-left">{{{key18}}}</td> <td class="infobox-right">{{{value18}}}</td> </tr> }}{{#if: {{{key19|}}} | <tr> <td class="infobox-left">{{{key19}}}</td> <td class="infobox-right">{{{value19}}}</td> </tr> }}{{#if: {{{key20|}}} | <tr> <td class="infobox-left">{{{key20}}}</td> <td class="infobox-right">{{{value20}}}</td> </tr> }} </table> </includeonly> <noinclude>This is the template for the robot infobox. Fields are: {{#if: {{{key1|}}} | * <code>{{{key1}}}</code> <code>{{{value1}}}</code> }} {{#if: {{{key2|}}} | * <code>{{{key2}}}</code> <code>{{{value2}}}</code> }} {{#if: {{{key3|}}} | * <code>{{{key3}}}</code> <code>{{{value3}}}</code> }} {{#if: {{{key4|}}} | * <code>{{{key4}}}</code> <code>{{{value4}}}</code> }} {{#if: {{{key5|}}} | * <code>{{{key5}}}</code> <code>{{{value5}}}</code> }} [[Category:Templates]]</noinclude> 766cbb3c5719538484dc22623466543525d8dab1 10 35 153 152 2024-04-24T07:19:46Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox"><tr><th colspan="2">{{{name}}}</th></tr>{{#if: {{{key1|}}} | <tr><td class="infobox-left">{{{key1}}}</td><td class="infobox-right">{{{value1}}}</td></tr> }}{{#if: {{{key2|}}} | <tr><td class="infobox-left">{{{key2}}}</td><td class="infobox-right">{{{value2}}}</td></tr> }}{{#if: {{{key3|}}} | <tr><td class="infobox-left">{{{key3}}}</td><td class="infobox-right">{{{value3}}}</td></tr> }}{{#if: {{{key4|}}} | <tr><td class="infobox-left">{{{key4}}}</td><td class="infobox-right">{{{value4}}}</td></tr> }}{{#if: {{{key5|}}} | <tr><td class="infobox-left">{{{key5}}}</td><td class="infobox-right">{{{value5}}}</td></tr> }}{{#if: {{{key6|}}} | <tr><td class="infobox-left">{{{key6}}}</td><td class="infobox-right">{{{value6}}}</td></tr> }}{{#if: {{{key7|}}} | <tr><td class="infobox-left">{{{key7}}}</td><td class="infobox-right">{{{value7}}}</td></tr> }}{{#if: {{{key8|}}} | <tr><td class="infobox-left">{{{key8}}}</td><td class="infobox-right">{{{value8}}}</td></tr> }}{{#if: {{{key9|}}} | <tr><td class="infobox-left">{{{key9}}}</td><td class="infobox-right">{{{value9}}}</td></tr> }}{{#if: {{{key10|}}} | <tr><td class="infobox-left">{{{key10}}}</td><td class="infobox-right">{{{value10}}}</td></tr> }}{{#if: {{{key11|}}} | <tr><td class="infobox-left">{{{key11}}}</td><td class="infobox-right">{{{value11}}}</td></tr> }}{{#if: {{{key12|}}} | <tr><td class="infobox-left">{{{key12}}}</td><td class="infobox-right">{{{value12}}}</td></tr> }}{{#if: {{{key13|}}} | <tr><td class="infobox-left">{{{key13}}}</td><td class="infobox-right">{{{value13}}}</td></tr> }}{{#if: {{{key14|}}} | <tr><td class="infobox-left">{{{key14}}}</td><td class="infobox-right">{{{value14}}}</td></tr> }}{{#if: {{{key15|}}} | <tr><td class="infobox-left">{{{key15}}}</td><td class="infobox-right">{{{value15}}}</td></tr> }}{{#if: {{{key16|}}} | <tr><td class="infobox-left">{{{key16}}}</td><td class="infobox-right">{{{value16}}}</td></tr> }}{{#if: {{{key17|}}} | <tr><td class="infobox-left">{{{key17}}}</td><td class="infobox-right">{{{value17}}}</td></tr> }}{{#if: {{{key18|}}} | <tr><td class="infobox-left">{{{key18}}}</td><td class="infobox-right">{{{value18}}}</td></tr> }}{{#if: {{{key19|}}} | <tr><td class="infobox-left">{{{key19}}}</td><td class="infobox-right">{{{value19}}}</td></tr> }}{{#if: {{{key20|}}} | <tr><td class="infobox-left">{{{key20}}}</td><td class="infobox-right">{{{value20}}}</td></tr> }}</table></includeonly> <noinclude>This is the template for the robot infobox. Fields are: {{#if: {{{key1|}}} | * <code>{{{key1}}}</code> <code>{{{value1}}}</code> }} {{#if: {{{key2|}}} | * <code>{{{key2}}}</code> <code>{{{value2}}}</code> }} {{#if: {{{key3|}}} | * <code>{{{key3}}}</code> <code>{{{value3}}}</code> }} {{#if: {{{key4|}}} | * <code>{{{key4}}}</code> <code>{{{value4}}}</code> }} {{#if: {{{key5|}}} | * <code>{{{key5}}}</code> <code>{{{value5}}}</code> }} [[Category:Templates]]</noinclude> 45102a63a1ca8ef42e14b016d0be3b81cf2a39f9 157 153 2024-04-24T07:25:29Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox"><tr><th colspan="2">{{{name}}}</th></tr>{{#if: {{{value1|}}} | <tr><td class="infobox-left">{{{key1}}}</td><td class="infobox-right">{{{value1}}}</td></tr> }}{{#if: {{{value2|}}} | <tr><td class="infobox-left">{{{key2}}}</td><td class="infobox-right">{{{value2}}}</td></tr> }}{{#if: {{{value3|}}} | <tr><td class="infobox-left">{{{key3}}}</td><td class="infobox-right">{{{value3}}}</td></tr> }}{{#if: {{{value4|}}} | <tr><td class="infobox-left">{{{key4}}}</td><td class="infobox-right">{{{value4}}}</td></tr> }}{{#if: {{{value5|}}} | <tr><td class="infobox-left">{{{key5}}}</td><td class="infobox-right">{{{value5}}}</td></tr> }}{{#if: {{{value6|}}} | <tr><td class="infobox-left">{{{key6}}}</td><td class="infobox-right">{{{value6}}}</td></tr> }}{{#if: {{{value7|}}} | <tr><td class="infobox-left">{{{key7}}}</td><td class="infobox-right">{{{value7}}}</td></tr> }}{{#if: {{{value8|}}} | <tr><td class="infobox-left">{{{key8}}}</td><td class="infobox-right">{{{value8}}}</td></tr> }}{{#if: {{{value9|}}} | <tr><td class="infobox-left">{{{key9}}}</td><td class="infobox-right">{{{value9}}}</td></tr> }}{{#if: {{{value10|}}} | <tr><td class="infobox-left">{{{key10}}}</td><td class="infobox-right">{{{value10}}}</td></tr> }}{{#if: {{{value11|}}} | <tr><td class="infobox-left">{{{key11}}}</td><td class="infobox-right">{{{value11}}}</td></tr> }}{{#if: {{{value12|}}} | <tr><td class="infobox-left">{{{key12}}}</td><td class="infobox-right">{{{value12}}}</td></tr> }}{{#if: {{{value13|}}} | <tr><td class="infobox-left">{{{key13}}}</td><td class="infobox-right">{{{value13}}}</td></tr> }}{{#if: {{{value14|}}} | <tr><td class="infobox-left">{{{key14}}}</td><td class="infobox-right">{{{value14}}}</td></tr> }}{{#if: {{{value15|}}} | <tr><td class="infobox-left">{{{key15}}}</td><td class="infobox-right">{{{value15}}}</td></tr> }}{{#if: {{{value16|}}} | <tr><td class="infobox-left">{{{key16}}}</td><td class="infobox-right">{{{value16}}}</td></tr> }}{{#if: {{{value17|}}} | <tr><td class="infobox-left">{{{key17}}}</td><td class="infobox-right">{{{value17}}}</td></tr> }}{{#if: {{{value18|}}} | <tr><td class="infobox-left">{{{key18}}}</td><td class="infobox-right">{{{value18}}}</td></tr> }}{{#if: {{{value19|}}} | <tr><td class="infobox-left">{{{key19}}}</td><td class="infobox-right">{{{value19}}}</td></tr> }}{{#if: {{{value20|}}} | <tr><td class="infobox-left">{{{key20}}}</td><td class="infobox-right">{{{value20}}}</td></tr> }}</table></includeonly> <noinclude>This is the template for the robot infobox. Fields are: {{#if: {{{value1|}}} | * <code>{{{key1}}}</code> <code>{{{value1}}}</code> }} {{#if: {{{value2|}}} | * <code>{{{key2}}}</code> <code>{{{value2}}}</code> }} {{#if: {{{value3|}}} | * <code>{{{key3}}}</code> <code>{{{value3}}}</code> }} {{#if: {{{value4|}}} | * <code>{{{key4}}}</code> <code>{{{value4}}}</code> }} {{#if: {{{value5|}}} | * <code>{{{key5}}}</code> <code>{{{value5}}}</code> }} [[Category:Templates]]</noinclude> aa6245087d4a18d29673c0b6fdaa5e450e0094fb 161 157 2024-04-24T07:27:23Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox"><tr><th colspan="2">{{{name}}}</th></tr>{{#if: {{{value1|}}} | <tr><td class="infobox-left">{{{key1}}}</td><td class="infobox-right">{{{value1}}}</td></tr> }}{{#if: {{{value2|}}} | <tr><td class="infobox-left">{{{key2}}}</td><td class="infobox-right">{{{value2}}}</td></tr> }}{{#if: {{{value3|}}} | <tr><td class="infobox-left">{{{key3}}}</td><td class="infobox-right">{{{value3}}}</td></tr> }}{{#if: {{{value4|}}} | <tr><td class="infobox-left">{{{key4}}}</td><td class="infobox-right">{{{value4}}}</td></tr> }}{{#if: {{{value5|}}} | <tr><td class="infobox-left">{{{key5}}}</td><td class="infobox-right">{{{value5}}}</td></tr> }}{{#if: {{{value6|}}} | <tr><td class="infobox-left">{{{key6}}}</td><td class="infobox-right">{{{value6}}}</td></tr> }}{{#if: {{{value7|}}} | <tr><td class="infobox-left">{{{key7}}}</td><td class="infobox-right">{{{value7}}}</td></tr> }}{{#if: {{{value8|}}} | <tr><td class="infobox-left">{{{key8}}}</td><td class="infobox-right">{{{value8}}}</td></tr> }}{{#if: {{{value9|}}} | <tr><td class="infobox-left">{{{key9}}}</td><td class="infobox-right">{{{value9}}}</td></tr> }}{{#if: {{{value10|}}} | <tr><td class="infobox-left">{{{key10}}}</td><td class="infobox-right">{{{value10}}}</td></tr> }}{{#if: {{{value11|}}} | <tr><td class="infobox-left">{{{key11}}}</td><td class="infobox-right">{{{value11}}}</td></tr> }}{{#if: {{{value12|}}} | <tr><td class="infobox-left">{{{key12}}}</td><td class="infobox-right">{{{value12}}}</td></tr> }}{{#if: {{{value13|}}} | <tr><td class="infobox-left">{{{key13}}}</td><td class="infobox-right">{{{value13}}}</td></tr> }}{{#if: {{{value14|}}} | <tr><td class="infobox-left">{{{key14}}}</td><td class="infobox-right">{{{value14}}}</td></tr> }}{{#if: {{{value15|}}} | <tr><td class="infobox-left">{{{key15}}}</td><td class="infobox-right">{{{value15}}}</td></tr> }}{{#if: {{{value16|}}} | <tr><td class="infobox-left">{{{key16}}}</td><td class="infobox-right">{{{value16}}}</td></tr> }}{{#if: {{{value17|}}} | <tr><td class="infobox-left">{{{key17}}}</td><td class="infobox-right">{{{value17}}}</td></tr> }}{{#if: {{{value18|}}} | <tr><td class="infobox-left">{{{key18}}}</td><td class="infobox-right">{{{value18}}}</td></tr> }}{{#if: {{{value19|}}} | <tr><td class="infobox-left">{{{key19}}}</td><td class="infobox-right">{{{value19}}}</td></tr> }}{{#if: {{{value20|}}} | <tr><td class="infobox-left">{{{key20}}}</td><td class="infobox-right">{{{value20}}}</td></tr> }}</table></includeonly> <noinclude>This is the template for the robot infobox. Fields are: {{#if: {{{key1|}}} | * <code>{{{key1}}}</code> <code>{{{value1}}}</code> }} {{#if: {{{key2|}}} | * <code>{{{key2}}}</code> <code>{{{value2}}}</code> }} {{#if: {{{key3|}}} | * <code>{{{key3}}}</code> <code>{{{value3}}}</code> }} {{#if: {{{key4|}}} | * <code>{{{key4}}}</code> <code>{{{value4}}}</code> }} {{#if: {{{key5|}}} | * <code>{{{key5}}}</code> <code>{{{value5}}}</code> }} [[Category:Templates]]</noinclude> f49f16982aa7cb7e0e2f9da1acc2ff7fd7798b50 184 161 2024-04-24T07:48:01Z Mrroboto 5 wikitext text/x-wiki <includeonly><table class="infobox"><tr><th colspan="2">{{{name}}}</th></tr>{{#if: {{{value1|}}} | <tr><td class="infobox-left">{{{key1}}}</td><td class="infobox-right">{{{value1}}}</td></tr> }}{{#if: {{{value2|}}} | <tr><td class="infobox-left">{{{key2}}}</td><td class="infobox-right">{{{value2}}}</td></tr> }}{{#if: {{{value3|}}} | <tr><td class="infobox-left">{{{key3}}}</td><td class="infobox-right">{{{value3}}}</td></tr> }}{{#if: {{{value4|}}} | <tr><td class="infobox-left">{{{key4}}}</td><td class="infobox-right">{{{value4}}}</td></tr> }}{{#if: {{{value5|}}} | <tr><td class="infobox-left">{{{key5}}}</td><td class="infobox-right">{{{value5}}}</td></tr> }}{{#if: {{{value6|}}} | <tr><td class="infobox-left">{{{key6}}}</td><td class="infobox-right">{{{value6}}}</td></tr> }}{{#if: {{{value7|}}} | <tr><td class="infobox-left">{{{key7}}}</td><td class="infobox-right">{{{value7}}}</td></tr> }}{{#if: {{{value8|}}} | <tr><td class="infobox-left">{{{key8}}}</td><td class="infobox-right">{{{value8}}}</td></tr> }}{{#if: {{{value9|}}} | <tr><td class="infobox-left">{{{key9}}}</td><td class="infobox-right">{{{value9}}}</td></tr> }}{{#if: {{{value10|}}} | <tr><td class="infobox-left">{{{key10}}}</td><td class="infobox-right">{{{value10}}}</td></tr> }}{{#if: {{{value11|}}} | <tr><td class="infobox-left">{{{key11}}}</td><td class="infobox-right">{{{value11}}}</td></tr> }}{{#if: {{{value12|}}} | <tr><td class="infobox-left">{{{key12}}}</td><td class="infobox-right">{{{value12}}}</td></tr> }}{{#if: {{{value13|}}} | <tr><td class="infobox-left">{{{key13}}}</td><td class="infobox-right">{{{value13}}}</td></tr> }}{{#if: {{{value14|}}} | <tr><td class="infobox-left">{{{key14}}}</td><td class="infobox-right">{{{value14}}}</td></tr> }}{{#if: {{{value15|}}} | <tr><td class="infobox-left">{{{key15}}}</td><td class="infobox-right">{{{value15}}}</td></tr> }}{{#if: {{{value16|}}} | <tr><td class="infobox-left">{{{key16}}}</td><td class="infobox-right">{{{value16}}}</td></tr> }}{{#if: {{{value17|}}} | <tr><td class="infobox-left">{{{key17}}}</td><td class="infobox-right">{{{value17}}}</td></tr> }}{{#if: {{{value18|}}} | <tr><td class="infobox-left">{{{key18}}}</td><td class="infobox-right">{{{value18}}}</td></tr> }}{{#if: {{{value19|}}} | <tr><td class="infobox-left">{{{key19}}}</td><td class="infobox-right">{{{value19}}}</td></tr> }}{{#if: {{{value20|}}} | <tr><td class="infobox-left">{{{key20}}}</td><td class="infobox-right">{{{value20}}}</td></tr> }}</table></includeonly><noinclude>This is the template for the robot infobox. Fields are: {{#if: {{{key1|}}} | * <code>{{{key1}}}</code> <code>{{{value1}}}</code> }} {{#if: {{{key2|}}} | * <code>{{{key2}}}</code> <code>{{{value2}}}</code> }} {{#if: {{{key3|}}} | * <code>{{{key3}}}</code> <code>{{{value3}}}</code> }} {{#if: {{{key4|}}} | * <code>{{{key4}}}</code> <code>{{{value4}}}</code> }} {{#if: {{{key5|}}} | * <code>{{{key5}}}</code> <code>{{{value5}}}</code> }} [[Category:Templates]]</noinclude> 7263756283434f53955e5728415cc65c7bbd6e6e 10 25 154 115 2024-04-24T07:21:38Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Country | value2 = {{{country}}} | key3 = {{#if: {{{website|}}} | Website }} | value3 = [{{{website}}} Website] | key4 = Robots | value4 = {{{robots}}} }} f2abf70039bcc3854452a5cdebaabcb05cf47ce5 159 154 2024-04-24T07:26:15Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Country | value2 = {{{country}}} | key3 = Website | value3 = {{#if: {{{website|}}} | [{{{website}}} Website] }} | key4 = Robots | value4 = {{{robots}}} }} 94f3e8738aa4a055425b603633a7e3bb631a08a5 160 159 2024-04-24T07:26:45Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name|}}} | key1 = Name | value1 = {{{name|}}} | key2 = Country | value2 = {{{country|}}} | key3 = Website | value3 = {{#if: {{{website|}}} | [{{{website}}} Website] }} | key4 = Robots | value4 ={{{robots|}}} }} 518ff8998c0efcf9fa4beb15cfc16f6b98e925db 162 160 2024-04-24T07:27:34Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name|}}} | key1 = Name | value1 = {{{name|}}} | key2 = Country | value2 = {{{country|}}} | key3 = Website | value3 = {{#if: {{{website|}}} | [{{{website}}} Website] }} | key4 = Robots | value4 = {{{robots|}}} }} 1a87279bf81ff1109842e4a9f846847a10d482ef 164 162 2024-04-24T07:28:27Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Country | value2 = {{{country|}}} | key3 = Website | value3 = {{#if: {{{website|}}} | [{{{website}}} Website] }} | key4 = Robots | value4 = {{{robots|}}} }} b795ceaeeb24837ae6e1d7c67313264ddde0fa74 170 164 2024-04-24T07:32:04Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Country | value2 = {{{country|}}} | key3 = Website | value3 = {{#if: {{{website_link|}}} | [{{{website_link}}} Website] }} | key4 = Robots | value4 = {{{robots|}}} }} f7f6f422d2fbac3594c45877b8a00efdf7533f17 10 28 155 151 2024-04-24T07:23:19Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization}}} | key3 = {{#if: {{{video|}}} | Video }} | value3 = [{{{video}}} Video] | key4 = Cost | value4 = {{{cost}}} | key5 = Height | value5 = {{{height}}} | key6 = Weight | value6 = {{{weight}}} | key7 = Lift Force | value7 = {{{lift_force}}} | key8 = {{#if: {{{purchase_link|}}} | Purchase }} | value8 = [{{{purchase_link}}} Link] }} 2297cc7b73ba1223b8d5ad7b78da82cc6d0878d6 158 155 2024-04-24T07:25:33Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization}}} | key3 = Video | value3 = {{#if: {{{video|}}} | [{{{video}}} Video] }} | key4 = Cost | value4 = {{{cost}}} | key5 = Height | value5 = {{{height}}} | key6 = Weight | value6 = {{{weight}}} | key7 = Lift Force | value7 = {{{lift_force}}} | key8 = Purchase | value8 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} }} f6c30963b2982e8060144b66ec27a71ae011798f 163 158 2024-04-24T07:28:20Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video|}}} | [{{{video}}} Video] }} | key4 = Cost | value4 = {{{cost|}}} | key5 = Height | value5 = {{{height|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = Lift Force | value7 = {{{lift_force|}}} | key8 = Purchase | value8 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} }} 17e56ad5d8a1656f1bbe9e70c495df453f7b3b0b 165 163 2024-04-24T07:29:36Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video|}}} | [{{{video}}} Video] }} | key4 = Cost | value4 = {{{cost|}}} | key5 = Height | value5 = {{{height|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = Lift Force | value7 = {{{lift_force|}}} | key8 = Battery Life | value8 = {{{battery_life|}}} | key8 = Purchase | value8 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} }} 034b28a570c64835ccc35995caefa2cf31731606 180 165 2024-04-24T07:42:41Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video_link|}}} | [{{{video_link}}} Video] }} | key4 = Cost | value4 = {{{cost|}}} | key5 = Height | value5 = {{{height|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = Lift Force | value7 = {{{lift_force|}}} | key8 = Battery Life | value8 = {{{battery_life|}}} | key8 = Purchase | value8 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} }} 228b40e88657101cad42969794b866563320727f 181 180 2024-04-24T07:42:57Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video_link|}}} | [{{{video_link}}} Video] }} | key4 = Cost | value4 = {{{cost|}}} | key5 = Height | value5 = {{{height|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = Lift Force | value7 = {{{lift_force|}}} | key8 = Battery Life | value8 = {{{battery_life|}}} | key9 = Purchase | value9 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} }} 8d7ee5633b31454ecde31f8058d00a371cbbf9fa 182 181 2024-04-24T07:45:38Z Mrroboto 5 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video_link|}}} | [{{{video_link}}} Video] }} | key4 = Cost | value4 = {{{cost|}}} | key5 = Height | value5 = {{{height|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = Speed | value7 = {{{speed|}}} | key8 = Lift Force | value8 = {{{lift_force|}}} | key9 = Battery Life | value9 = {{{battery_life|}}} | key10 = Battery Capacity | value10 = {{{battery_capacity|}}} | key11 = Purchase | value11 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key12 = Number Made | value12 = {{{number_made|}}} | key13 = DoF | value13 = {{{dof|}}} | key14 = Status | value14 = {{{status|}}} }} fed766ff398d58d0b86ec28c3ae5bca9a62c414a 0 22 156 150 2024-04-24T07:23:42Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] aec3b64acc1be73c94c3e1cc4527fa30c10f908c 167 156 2024-04-24T07:30:58Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 931a6a4042eff44ffac49b42e7a93565d13b34ea 178 167 2024-04-24T07:40:04Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} Tesla began work on the Optimus robot in 2021. [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 7d6acc73a5cc062cf4c5b7ac89f85af8aafdbe60 179 178 2024-04-24T07:40:11Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} Tesla began work on the Optimus robot in 2021. [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 0870360f622a742982fc117229041450cba128ee 194 179 2024-04-24T07:54:34Z Mrroboto 5 wikitext text/x-wiki Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} Tesla began work on the Optimus robot in 2021. [[File:Optimus Tesla (1).jpg|none|300px|The Tesla Optimus on display|thumb]] [[Category:Robots]] 7d6acc73a5cc062cf4c5b7ac89f85af8aafdbe60 11 36 166 2024-04-24T07:30:44Z Mrroboto 5 Created page with "{{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video_link|}}} | [{{..." wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video_link|}}} | [{{{video_link}}} Video] }} | key4 = Cost | value4 = {{{cost|}}} | key5 = Height | value5 = {{{height|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = Lift Force | value7 = {{{lift_force|}}} | key8 = Battery Life | value8 = {{{battery_life|}}} | key8 = Purchase | value8 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} }} Fields: * <code>organization</code> * <code>video_link</code> * <code>cost</code> * <code>height</code> * <code>weight</code> * <code>lift_force</code> * <code>battery_life</code> * <code>purchase_link</code> b58f783908efc95ee8d904c149fc94db5a4cb350 169 166 2024-04-24T07:31:55Z Mrroboto 5 wikitext text/x-wiki Fields: * <code>organization</code> * <code>video_link</code> * <code>cost</code> * <code>height</code> * <code>weight</code> * <code>lift_force</code> * <code>battery_life</code> * <code>purchase_link</code> 15786e5eda391575690876877958896df4bc8b29 11 37 168 2024-04-24T07:31:42Z Mrroboto 5 Created page with "Fields: * <code>name</code> * <code>country</code> * <code>website_link</code> * <code>robots</code>" wikitext text/x-wiki Fields: * <code>name</code> * <code>country</code> * <code>website_link</code> * <code>robots</code> 52f848b4c5c83e9a0cef937520d89ac8700b0216 0 7 171 122 2024-04-24T07:32:12Z Mrroboto 5 wikitext text/x-wiki Tesla is building a humanoid robot called [[Optimus]]. {{infobox company | name = Tesla | country = United States | website_link = https://www.tesla.com/ | robots = [[Optimus]] }} [[Category:Companies]] f7819b06a63448554158caeb318c60621ac3e1c8 8 27 172 142 2024-04-24T07:34:24Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox td .infobox-left { background-color: #e0e0e0; text-align: center; padding: 2px 5px; text-align: center; background-color: #cccccc; } .infobox td { padding: 2px 5px; } 7226ebadb8dc8ec772d1c99745ed12a54d85e8c3 173 172 2024-04-24T07:34:31Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox td .infobox-left { background-color: #e0e0e0; text-align: center; padding: 2px 5px; } .infobox td { padding: 2px 5px; } d94bac92d8b6151913f78c145149d50d400beb0c 174 173 2024-04-24T07:34:45Z Admin 1 Undo revision 173 by [[Special:Contributions/Admin|Admin]] ([[User talk:Admin|talk]]) css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox td .infobox-left { background-color: #e0e0e0; text-align: center; padding: 2px 5px; text-align: center; background-color: #cccccc; } .infobox td { padding: 2px 5px; } 7226ebadb8dc8ec772d1c99745ed12a54d85e8c3 175 174 2024-04-24T07:35:07Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox th { background-color: #e0e0e0; padding: 2px 5px; text-align: center; } .infobox td { padding: 2px 5px; } ab5bbf2daca6bff8d8e124f4aea5f663ab05a507 176 175 2024-04-24T07:35:41Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox th { background-color: #e0e0e0; padding: 2px 5px; text-align: center; } .infobox td { padding: 2px 5px; } .infobox td .infobox-left { font-style: italic; } f1df76d6752a57a3c0a9bbd500fc9f237cf1687b 177 176 2024-04-24T07:36:42Z Admin 1 css text/css /* CSS placed here will be applied to all skins */ .infobox { border: 1px solid #aaaaaa; background-color: #f9f9f9; padding: 5px; width: 300px; font-size: 90%; margin-top: 20px; margin-bottom: 20px; } .infobox th { background-color: #e0e0e0; padding: 2px 5px; text-align: center; } .infobox td { padding: 2px 5px; } .infobox-left { font-style: italic; } 36098d154d5c7b8163de04b23dd304ea0d17fc59 0 24 183 89 2024-04-24T07:47:41Z Mrroboto 5 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by Agility Robotics. {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. eced16f955f7717f14459f0883b981c6e3a40efb 186 183 2024-04-24T07:50:01Z Mrroboto 5 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by Agility Robotics. {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." [[File:Cassie.jpg|thumb]] ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. 8ac154ab6974f0be85312efe9de03c5089394414 187 186 2024-04-24T07:50:13Z Mrroboto 5 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by Agility Robotics. {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} [[File:Cassie.jpg|thumb]] ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. cc014401b4c3e47a5890517e194d56b4dec924e5 188 187 2024-04-24T07:50:26Z Mrroboto 5 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility Robotics]]. [[File:Cassie.jpg|thumb]] {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. 3d070e6b71ed9b1faf11a4c08d11efe452d63228 190 188 2024-04-24T07:50:47Z Mrroboto 5 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility]]. [[File:Cassie.jpg|thumb]] {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. 6a238ec32dfd5172595f23d0ea9615af3bc37401 191 190 2024-04-24T07:51:27Z Mrroboto 5 wikitext text/x-wiki Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility]]. [[File:Cassie.jpg|right|200px|thumb]] {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. 916110f5eed226049da63b407645cb2008fc2c1e 202 191 2024-04-24T07:59:57Z Mrroboto 5 wikitext text/x-wiki [[File:Cassie.jpg|right|200px|thumb]] Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility]]. {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. e15af65042f9102989b6544fbbad7470c9ad177c 6 38 185 2024-04-24T07:49:51Z Mrroboto 5 wikitext text/x-wiki A view of the Cassie robot standing 311a30afec019916685992a076131ea91dfd145c 0 8 189 40 2024-04-24T07:50:39Z Mrroboto 5 wikitext text/x-wiki Agility has built several robots. Their humanoid robot is called [[Digit]]. [[Category:Companies]] b7c7496933b1b37a0232cd595feac0d64b1c1dd8 192 189 2024-04-24T07:52:22Z Mrroboto 5 wikitext text/x-wiki Agility has built several robots. Their humanoid robot is called [[Digit]]. {{infobox company | name = Agility | country = United States | website_link = https://agilityrobotics.com/ | robots = [[Cassie]], [[Digit]] }} [[Category:Companies]] 18dbe0ca7deb2aa4672228d807ad4133dd44b880 0 1 193 123 2024-04-24T07:52:44Z Mrroboto 5 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 51c42062c13c144063aab44be8cfe2166803dc5e 6 39 195 2024-04-24T07:55:34Z Mrroboto 5 wikitext text/x-wiki Stompy standing up 5472b34d182146bd2bda9716569528d1b3a7939e 0 2 196 46 2024-04-24T07:56:36Z Mrroboto 5 wikitext text/x-wiki [[File:Stompy.jpg|right|200px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = $10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robots]] 23514b97861e631b335e4c3af26ff3bd9bad453b 197 196 2024-04-24T07:56:57Z Mrroboto 5 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = $10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions [[Category:Robots]] c5629ea3dd2720c376229c77117ad7afdaa42b14 198 197 2024-04-24T07:57:52Z Mrroboto 5 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = $10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] 3ce03b40d040a8253e96804412f92267b0b94b26 0 5 199 48 2024-04-24T07:58:24Z Mrroboto 5 wikitext text/x-wiki [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} [[Category:Companies]] 43baae20de1554a7100d463518d11ccd93034f9f 201 199 2024-04-24T07:59:44Z Mrroboto 5 wikitext text/x-wiki [[File:Logo.png|right|200px|thumb]] [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} [[Category:Companies]] 2bfd4249601bb542b3fac0762535aa3816d07a4d 6 40 200 2024-04-24T07:59:02Z Mrroboto 5 wikitext text/x-wiki The K-Scale Labs logo ba44c02f5ecb30e4a9204639ba9e9fbe1d1bed2b 0 41 203 2024-04-24T08:05:21Z Mrroboto 5 Created page with "Walk-through and notes regarding how to design and ship a PCB." wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. b36ceecece334f4d82f0be420552f11918204ac4 204 203 2024-04-24T08:07:37Z Mrroboto 5 wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] 7f7f8a58a029adbd3a13d489a3f69713936315cb 14 42 205 2024-04-24T08:07:43Z Mrroboto 5 Created blank page wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 206 205 2024-04-24T08:07:59Z Mrroboto 5 wikitext text/x-wiki Pages related to building hardware. 766f95387b6b8e9a66a15fa7d341a630d19ef771 14 43 207 2024-04-24T08:08:28Z Mrroboto 5 Created page with "Pages designated as guides." wikitext text/x-wiki Pages designated as guides. de810d1cf8dd81ed0f0d0f99f2f40c7b98729347 14 44 208 2024-04-24T08:08:40Z Mrroboto 5 Created page with "Pages related to dealing with electronics." wikitext text/x-wiki Pages related to dealing with electronics. 611b344bd910427314f2c0b6ba39fd3d1c308d21 0 1 209 193 2024-04-24T08:10:17Z Mrroboto 5 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[DigitV3]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 9df948bf81e237039fd5847a889c089341c11fec 220 209 2024-04-24T08:21:00Z Mrroboto 5 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[EVE]], [[NEO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 49045503b3bf433f08ab8827d3fe5aca5ab84a78 221 220 2024-04-24T08:21:13Z Mrroboto 5 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} ab3ca86908b06d6faef26f8fcb5494cec1095dca 235 221 2024-04-24T09:14:22Z Mrroboto 5 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides|Guides]] | Category for pages which act as guides |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} ee04bb47ccb82de0f53870d06bed059cae75b3e3 237 235 2024-04-24T09:15:23Z Mrroboto 5 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides|Guides]] | Category for pages which act as guides |- | [[:Category:Electronics|Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware|Hardware]] | Category for pages relating to hardware |- | [[:Category:Software|Software]] | Category for pages relating to software |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 3c5cc1f27d8f48d3883d8abdf56020a79cb82490 238 237 2024-04-24T09:15:35Z Mrroboto 5 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 338c4b364389cf4d8d4a78b5155c627387cbe867 0 45 210 2024-04-24T08:11:02Z Mrroboto 5 Created page with "Build guide for constructing [[Stompy]]. [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]]" wikitext text/x-wiki Build guide for constructing [[Stompy]]. [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] a08d7afa5a13fb8db1968afed18fda915a14ae5d 0 2 211 198 2024-04-24T08:13:29Z Mrroboto 5 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = $10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. See the [[Stompy Build Guide|build guide]] for a walk-through of how to build one yourself. = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] cddc52a4487dd2ca873b8977afbfa8cb0322f7df 0 24 212 202 2024-04-24T08:14:28Z Mrroboto 5 wikitext text/x-wiki [[File:Cassie.jpg|right|200px|thumb]] Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility]]. {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = $250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. [[Category: Robots]] 34abd9c527bfa7f653060c905b52d1f576e277cf 0 3 213 75 2024-04-24T08:15:32Z Mrroboto 5 wikitext text/x-wiki It is available for purchase [https://shop.unitree.com/products/unitree-h1 here]. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = $150,000 | purchase_link = https://shop.unitree.com/products/unitree-h1 }} [[Category: Robots]] d3f470d790d94fc2fa4a6b3611b62215d4052ed4 0 22 214 194 2024-04-24T08:16:58Z Mrroboto 5 wikitext text/x-wiki [[File:Optimus Tesla (1).jpg|right|200px|thumb]] Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} Tesla began work on the Optimus robot in 2021. [[Category:Robots]] b35d08f8818c2b84912702c2888bf2247d48a8e8 236 214 2024-04-24T09:14:49Z User2024 6 wikitext text/x-wiki [[File:Optimus Tesla (1).jpg|right|200px|thumb]] Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | height = 5 ft 8 in (173 cm) | weight = 58 kg | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} Tesla began work on the Optimus robot in 2021. [[Category:Robots]] 4ebfc66ae0ce2466c9f41ce3eb73732822bcd6ad 0 9 215 39 2024-04-24T08:17:45Z Mrroboto 5 wikitext text/x-wiki Sanctuary AI is a humanoid robot company. Their robot is called [[Phoenix]]. {{infobox company | name = Sanctuary | country = United States | website_link = https://sanctuary.ai/ | robots = [[Phoenix]] }} [[Category:Companies]] 73c4775e3cfc7f680e9efb6e343f7d807c4c2146 0 10 216 74 2024-04-24T08:18:46Z Mrroboto 5 wikitext text/x-wiki [https://www.1x.tech/ 1X] (formerly known as Halodi Robotics) is a humanoid robotics company based in Moss, Norway. They have two robots: [[Eve]] and [[Neo]]. [[Eve]] is a wheeled robot while [[Neo]] has legs. The company is known for it's high torque BLDC motors that they developed in house. Those BLDC motors are paired with low gear ratio cable drives. Eve and Neo are designed for safe human interaction by reducing actuator inertia. {{infobox company | name = 1X Technologies | country = United States | website_link = https://www.1x.tech/ | robots = [[Eve]], [[Neo]] }} [[Category:Companies]] 5a8403a4c8da4adcdbece6263ec1a8be7d0dd465 0 6 217 37 2024-04-24T08:19:27Z Mrroboto 5 wikitext text/x-wiki Unitree is a company based out of China which has built a number of different types of robots. {{infobox company | name = Unitree | country = China | website_link = https://www.unitree.com/ | robots = [[H1]] }} [[Category:Companies]] 73d68bb511ac8eaa83cfca4764c1f399c20f781e 0 12 218 43 2024-04-24T08:19:59Z Mrroboto 5 wikitext text/x-wiki [https://physicalintelligence.company/ Physical Intelligence] is a company based in the Bay Area which is building foundation models for embodied AI. {{infobox company | name = Physical Intelligence | country = United States | website_link = https://physicalintelligence.company/ }} [[Category:Companies]] 14c75bc8d5e6d8e6935a25d73db01c8c448760ed 0 11 219 44 2024-04-24T08:20:23Z Mrroboto 5 wikitext text/x-wiki Skild is a stealth foundation model startup started by two faculty members from Carnegie Mellon University. {{infobox company | name = Skild | country = United States }} === Articles === * [https://www.theinformation.com/articles/venture-fomo-hits-robotics-as-young-startup-gets-1-5-billion-valuation Venture FOMO Hits Robotics as Young Startup Gets $1.5 Billion Valuation] [[Category:Companies]] 880bcffd8281a04a0d81d27cab54ef044e348e99 0 20 222 59 2024-04-24T09:01:46Z Mrroboto 5 wikitext text/x-wiki The Jetson Orin is a development board from Nvidia. === CAN Bus === See [https://docs.nvidia.com/jetson/archives/r34.1/DeveloperGuide/text/HR/ControllerAreaNetworkCan.html here] for notes on configuring the CAN bus for the Jetson. Install dependencies: <syntaxhighlight lang="bash"> sudo add-apt-repository ppa:deadsnakes/ppa sudo apt update sudo apt upgrade sudo apt install g++ python3.11-dev </syntaxhighlight> Initialize the CAN bus on startup: <syntaxhighlight lang="bash"> #!/bin/bash # Set pinmux. busybox devmem 0x0c303000 32 0x0000C400 busybox devmem 0x0c303008 32 0x0000C458 busybox devmem 0x0c303010 32 0x0000C400 busybox devmem 0x0c303018 32 0x0000C458 # Install modules. modprobe can modprobe can_raw modprobe mttcan # Turn off CAN. ip link set down can0 ip link set down can1 # Set parameters. ip link set can0 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off ip link set can1 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off # Turn on CAN. ip link set up can0 ip link set up can1 </syntaxhighlight> You can run this script automatically on startup by writing a service configuration to (for example) <code>/etc/systemd/system/can_setup.service</code> <syntaxhighlight lang="text"> [Unit] Description=Initialize CAN Interfaces After=network.target [Service] Type=oneshot ExecStart=/opt/kscale/enable_can.sh RemainAfterExit=true [Install] WantedBy=multi-user.target </syntaxhighlight> To enable this, run: <syntaxhighlight lang="bash"> sudo systemctl enable can_setup sudo systemctl start can_setup </syntaxhighlight> === Cameras === ==== Arducam IMX 219 ==== * [https://www.arducam.com/product/arducam-imx219-multi-camera-kit-for-the-nvidia-jetson-agx-orin/ Product Page] ** Shipping was pretty fast ** Order a couple backup cameras because a couple of the cameras that they shipped came busted * [https://docs.arducam.com/Nvidia-Jetson-Camera/Nvidia-Jetson-Orin-Series/NVIDIA-Jetson-AGX-Orin/Quick-Start-Guide/ Quick start guide] Run the installation script: <syntaxhighlight lang="bash"> wget https://github.com/ArduCAM/MIPI_Camera/releases/download/v0.0.3/install_full.sh chmod u+x install_full.sh ./install_full.sh -m imx219 </syntaxhighlight> Supported kernel versions (see releases [https://github.com/ArduCAM/MIPI_Camera/releases here]): * <code>5.10.104-tegra-35.3.1</code> * <code>5.10.120-tegra-35.4.1</code> Install an older kernel from [https://developer.nvidia.com/embedded/jetson-linux-archive here]. This required downgrading to Ubuntu 20.04 (only changing <code>/etc/os-version</code>). Install dependencies: <syntaxhighlight lang="bash"> sudo apt update sudo apt install \ gstreamer1.0-tools \ gstreamer1.0-alsa \ gstreamer1.0-plugins-base \ gstreamer1.0-plugins-good \ gstreamer1.0-plugins-bad \ gstreamer1.0-plugins-ugly \ gstreamer1.0-libav sudo apt install libgstreamer1.0-dev \ libgstreamer-plugins-base1.0-dev \ libgstreamer-plugins-good1.0-dev \ libgstreamer-plugins-bad1.0-dev sudo apt install \ v4l-utils \ ffmpeg </syntaxhighlight> Make sure the camera shows up: <syntaxhighlight lang="bash"> v4l2-ctl --list-formats-ext </syntaxhighlight> Capture a frame from the camera: <syntaxhighlight lang="bash"> gst-launch-1.0 nvarguscamerasrc sensor-id=0 ! "video/x-raw(memory:NVMM), width=1280, height=720, framerate=60/1" ! nvvidconv ! jpegenc snapshot=TRUE ! filesink location=test.jpg </syntaxhighlight> Alternatively, use the following Python code: <syntaxhighlight lang="bash"> import cv2 gst_str = ( 'nvarguscamerasrc sensor-id=0 ! ' 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)720, format=(string)NV12, framerate=(fraction)60/1 ! ' 'nvvidconv flip-method=0 ! ' 'video/x-raw, width=(int)1280, height=(int)720, format=(string)BGRx ! ' 'videoconvert ! ' 'video/x-raw, format=(string)BGR ! ' 'appsink' ) cap = cv2.VideoCapture(gst_str, cv2.CAP_GSTREAMER) while True: ret, frame = cap.read() if ret: print(frame) if cv2.waitKey(1) & 0xFF == ord('q'): break else: break cap.release() cv2.destroyAllWindows() </syntaxhighlight> === IMU === We're using the [https://ozzmaker.com/product/berryimu-accelerometer-gyroscope-magnetometer-barometricaltitude-sensor/ BerryIMU v3]. To use it, connect pin 3 on the Jetson to SDA and pin 5 to SCL for I2C bus 7. You can verify the connection is successful if the following command matches: <syntaxhighlight lang="bash"> $ sudo i2cdetect -y -r 7 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- 1c -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- 6a -- -- -- -- -- 70: -- -- -- -- -- -- -- 77 </syntaxhighlight> The equivalent command on the Raspberry Pi should use bus 1: <syntaxhighlight lang="bash"> sudo i2cdetect -y -r 1 </syntaxhighlight> The default addresses are: * <code>0x6A</code>: Gyroscope and accelerometer * <code>0x1C</code>: Magnetometer * <code>0x77</code>: Barometer [[Category: Hardware]] [[Category: Electronics]] 7021fbaed1011dfc8b981f679889c12d097b78a8 225 222 2024-04-24T09:04:56Z Mrroboto 5 wikitext text/x-wiki The Jetson Orin is a development board from Nvidia. === CAN Bus === See [https://docs.nvidia.com/jetson/archives/r34.1/DeveloperGuide/text/HR/ControllerAreaNetworkCan.html here] for notes on configuring the CAN bus for the Jetson. [[File:Can bus connections 2.png|none|200px|thumb]] Install dependencies: <syntaxhighlight lang="bash"> sudo add-apt-repository ppa:deadsnakes/ppa sudo apt update sudo apt upgrade sudo apt install g++ python3.11-dev </syntaxhighlight> Initialize the CAN bus on startup: <syntaxhighlight lang="bash"> #!/bin/bash # Set pinmux. busybox devmem 0x0c303000 32 0x0000C400 busybox devmem 0x0c303008 32 0x0000C458 busybox devmem 0x0c303010 32 0x0000C400 busybox devmem 0x0c303018 32 0x0000C458 # Install modules. modprobe can modprobe can_raw modprobe mttcan # Turn off CAN. ip link set down can0 ip link set down can1 # Set parameters. ip link set can0 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off ip link set can1 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off # Turn on CAN. ip link set up can0 ip link set up can1 </syntaxhighlight> You can run this script automatically on startup by writing a service configuration to (for example) <code>/etc/systemd/system/can_setup.service</code> <syntaxhighlight lang="text"> [Unit] Description=Initialize CAN Interfaces After=network.target [Service] Type=oneshot ExecStart=/opt/kscale/enable_can.sh RemainAfterExit=true [Install] WantedBy=multi-user.target </syntaxhighlight> To enable this, run: <syntaxhighlight lang="bash"> sudo systemctl enable can_setup sudo systemctl start can_setup </syntaxhighlight> === Cameras === ==== Arducam IMX 219 ==== * [https://www.arducam.com/product/arducam-imx219-multi-camera-kit-for-the-nvidia-jetson-agx-orin/ Product Page] ** Shipping was pretty fast ** Order a couple backup cameras because a couple of the cameras that they shipped came busted * [https://docs.arducam.com/Nvidia-Jetson-Camera/Nvidia-Jetson-Orin-Series/NVIDIA-Jetson-AGX-Orin/Quick-Start-Guide/ Quick start guide] Run the installation script: <syntaxhighlight lang="bash"> wget https://github.com/ArduCAM/MIPI_Camera/releases/download/v0.0.3/install_full.sh chmod u+x install_full.sh ./install_full.sh -m imx219 </syntaxhighlight> Supported kernel versions (see releases [https://github.com/ArduCAM/MIPI_Camera/releases here]): * <code>5.10.104-tegra-35.3.1</code> * <code>5.10.120-tegra-35.4.1</code> Install an older kernel from [https://developer.nvidia.com/embedded/jetson-linux-archive here]. This required downgrading to Ubuntu 20.04 (only changing <code>/etc/os-version</code>). Install dependencies: <syntaxhighlight lang="bash"> sudo apt update sudo apt install \ gstreamer1.0-tools \ gstreamer1.0-alsa \ gstreamer1.0-plugins-base \ gstreamer1.0-plugins-good \ gstreamer1.0-plugins-bad \ gstreamer1.0-plugins-ugly \ gstreamer1.0-libav sudo apt install libgstreamer1.0-dev \ libgstreamer-plugins-base1.0-dev \ libgstreamer-plugins-good1.0-dev \ libgstreamer-plugins-bad1.0-dev sudo apt install \ v4l-utils \ ffmpeg </syntaxhighlight> Make sure the camera shows up: <syntaxhighlight lang="bash"> v4l2-ctl --list-formats-ext </syntaxhighlight> Capture a frame from the camera: <syntaxhighlight lang="bash"> gst-launch-1.0 nvarguscamerasrc sensor-id=0 ! "video/x-raw(memory:NVMM), width=1280, height=720, framerate=60/1" ! nvvidconv ! jpegenc snapshot=TRUE ! filesink location=test.jpg </syntaxhighlight> Alternatively, use the following Python code: <syntaxhighlight lang="bash"> import cv2 gst_str = ( 'nvarguscamerasrc sensor-id=0 ! ' 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)720, format=(string)NV12, framerate=(fraction)60/1 ! ' 'nvvidconv flip-method=0 ! ' 'video/x-raw, width=(int)1280, height=(int)720, format=(string)BGRx ! ' 'videoconvert ! ' 'video/x-raw, format=(string)BGR ! ' 'appsink' ) cap = cv2.VideoCapture(gst_str, cv2.CAP_GSTREAMER) while True: ret, frame = cap.read() if ret: print(frame) if cv2.waitKey(1) & 0xFF == ord('q'): break else: break cap.release() cv2.destroyAllWindows() </syntaxhighlight> === IMU === We're using the [https://ozzmaker.com/product/berryimu-accelerometer-gyroscope-magnetometer-barometricaltitude-sensor/ BerryIMU v3]. To use it, connect pin 3 on the Jetson to SDA and pin 5 to SCL for I2C bus 7. You can verify the connection is successful if the following command matches: <syntaxhighlight lang="bash"> $ sudo i2cdetect -y -r 7 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- 1c -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- 6a -- -- -- -- -- 70: -- -- -- -- -- -- -- 77 </syntaxhighlight> The equivalent command on the Raspberry Pi should use bus 1: <syntaxhighlight lang="bash"> sudo i2cdetect -y -r 1 </syntaxhighlight> The default addresses are: * <code>0x6A</code>: Gyroscope and accelerometer * <code>0x1C</code>: Magnetometer * <code>0x77</code>: Barometer [[Category: Hardware]] [[Category: Electronics]] 51030b57b98b338a704c3f6daae5082cb8b122c3 6 46 223 2024-04-24T09:02:43Z Mrroboto 5 wikitext text/x-wiki Image of the correctly-connected CAN busses 32b09284eeba2cda52c2fa7a0b317a199e33897c 6 47 224 2024-04-24T09:04:28Z Mrroboto 5 wikitext text/x-wiki Image of the CAN bus connections on the Jetson c2dc41b9f357149d1858ca42a66c4455f810a152 6 48 226 2024-04-24T09:06:07Z Mrroboto 5 wikitext text/x-wiki Diagram showing the Project Aria extrinsics 04e5a7eaac48013ebad9eae62371b20dc0622dbc 0 21 227 64 2024-04-24T09:06:10Z Mrroboto 5 wikitext text/x-wiki AR glasses for data capture from Meta. === References === ==== Links ==== * [https://www.projectaria.com/ Website] * [https://docs.ego-exo4d-data.org/ Ego4D Documentation] * [https://facebookresearch.github.io/projectaria_tools/docs/intro Project Aria Documentation] ** [https://facebookresearch.github.io/projectaria_tools/docs/data_formats/mps/mps_trajectory Specific page regarding trajectories] ==== Datasets ==== * [https://www.projectaria.com/datasets/apd/ APD Dataset] === Extrinsics === Here is a diagram showing the extrinsics of the various cameras on the Project Aria headset. [[File:Project aria extrinsics.png|thumb]] 2e546dbc3c7780e4f8480b2e83e53f078f31fe95 228 227 2024-04-24T09:06:20Z Mrroboto 5 wikitext text/x-wiki AR glasses for data capture from Meta. === References === ==== Links ==== * [https://www.projectaria.com/ Website] * [https://docs.ego-exo4d-data.org/ Ego4D Documentation] * [https://facebookresearch.github.io/projectaria_tools/docs/intro Project Aria Documentation] ** [https://facebookresearch.github.io/projectaria_tools/docs/data_formats/mps/mps_trajectory Specific page regarding trajectories] ==== Datasets ==== * [https://www.projectaria.com/datasets/apd/ APD Dataset] === Extrinsics === Here is a diagram showing the extrinsics of the various cameras on the Project Aria headset. [[File:Project aria extrinsics.png|none|thumb]] 1838dd0211ffcc31ab2851c982c90f64b2da43d0 229 228 2024-04-24T09:06:28Z Mrroboto 5 wikitext text/x-wiki AR glasses for data capture from Meta. === References === ==== Links ==== * [https://www.projectaria.com/ Website] * [https://docs.ego-exo4d-data.org/ Ego4D Documentation] * [https://facebookresearch.github.io/projectaria_tools/docs/intro Project Aria Documentation] ** [https://facebookresearch.github.io/projectaria_tools/docs/data_formats/mps/mps_trajectory Specific page regarding trajectories] ==== Datasets ==== * [https://www.projectaria.com/datasets/apd/ APD Dataset] === Extrinsics === Here is a diagram showing the extrinsics of the various cameras on the Project Aria headset. [[File:Project aria extrinsics.png|none|400px|thumb]] 6581aaecf783619297e517f9505b51a22722e639 0 49 230 2024-04-24T09:08:53Z Mrroboto 5 Created page with "Serial Peripheral Interface (SPI) is commonly used for connecting to peripheral devices. === Conventions === * <code>CS</code> is Chip Select ** On Raspberry Pi, this is <co..." wikitext text/x-wiki Serial Peripheral Interface (SPI) is commonly used for connecting to peripheral devices. === Conventions === * <code>CS</code> is Chip Select ** On Raspberry Pi, this is <code>CE0</code> or <code>CE1</code> ** This is a digital signal that tells the slave device to listen to the master * <code>DC</code> is Data/Command ** This is a digital signal that tells the slave device whether the data on the <code>MOSI</code> line is a command or data * <code>SDA</code> is data line ** Also called <code>MOSI</code> (Master Out Slave In) or <code>DIN</code> (Data In) ** This is the line that the master sends data to the slave * <code>SCL</code> is clock line ** Also called <code>CLK</code> or <code>SCLK</code> (Serial Clock) ** This is the line that the master uses to send clock pulses to the slave * <code>RST</code> is reset ** This is a digital signal that resets the slave device e7227546fb1ca6b3861518cf846bc4912ea61391 231 230 2024-04-24T09:09:44Z Mrroboto 5 wikitext text/x-wiki Serial Peripheral Interface (SPI) is commonly used for connecting to peripheral devices. A commonly used alternative is [[Inter-Integrated Circuit (I2C)]]. === Conventions === * <code>CS</code> is Chip Select ** On Raspberry Pi, this is <code>CE0</code> or <code>CE1</code> ** This is a digital signal that tells the slave device to listen to the master * <code>DC</code> is Data/Command ** This is a digital signal that tells the slave device whether the data on the <code>MOSI</code> line is a command or data * <code>SDA</code> is data line ** Also called <code>MOSI</code> (Master Out Slave In) or <code>DIN</code> (Data In) ** This is the line that the master sends data to the slave * <code>SCL</code> is clock line ** Also called <code>CLK</code> or <code>SCLK</code> (Serial Clock) ** This is the line that the master uses to send clock pulses to the slave * <code>RST</code> is reset ** This is a digital signal that resets the slave device [[Category: Communication]] bff26dc83f2c170b69f533a368651e6349ef9790 0 50 232 2024-04-24T09:11:22Z Mrroboto 5 Created page with "=== Characteristics === * Bi-directional * Widely used protocol for short-distance communication * <code>SDA</code> and <code>SCL</code> ** Both are pulled high Category:..." wikitext text/x-wiki === Characteristics === * Bi-directional * Widely used protocol for short-distance communication * <code>SDA</code> and <code>SCL</code> ** Both are pulled high [[Category: Communication]] 739be617979997d5eccf033ae317abb2ed9b6696 233 232 2024-04-24T09:11:57Z Mrroboto 5 wikitext text/x-wiki Inter-Integrated Circuit (I2C) is commonly used for connecting to peripheral devices. A commonly used alternative is [[Serial Peripheral Interface (SPI)]]. === Characteristics === * Bi-directional * Widely used protocol for short-distance communication * <code>SDA</code> and <code>SCL</code> ** Both are pulled high [[Category: Communication]] 12297408c799ce1d58e4cfdfff520a06042fb29b 14 51 234 2024-04-24T09:13:27Z Mrroboto 5 Created page with "Pages relating to various communication protocols." wikitext text/x-wiki Pages relating to various communication protocols. 12975d0a5247d36763f2e885c4ad0ec206a88e07 14 52 239 2024-04-24T09:15:48Z Mrroboto 5 Created page with "Pages which have something to do with software." wikitext text/x-wiki Pages which have something to do with software. 240de77297915a3a3e5c6fd7fe9a691cb7984478 0 32 240 132 2024-04-24T09:15:59Z Mrroboto 5 wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks ===Isaac Gym=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===Gymnasium=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. == Training algorithms == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] a3b43db2638b0e61cc57b04ad62aab6a73999c1f 0 53 241 2024-04-24T10:05:25Z User2024 6 Created page with "Phoenix is a humanoid robot from [[[https://sanctuary.ai/ Sanctuary AI]]]. {{infobox robot | name = Phoenix | organization = [[Sanctuary AI]] | height = 170 cm | weight = 70..." wikitext text/x-wiki Phoenix is a humanoid robot from [[[https://sanctuary.ai/ Sanctuary AI]]]. {{infobox robot | name = Phoenix | organization = [[Sanctuary AI]] | height = 170 cm | weight = 70 kg | two_hand_payload = 25 }} [[Category:Robots]] b17b68e2833f6b769574646edc48589265341950 242 241 2024-04-24T10:06:41Z User2024 6 wikitext text/x-wiki Phoenix is a humanoid robot from [https://sanctuary.ai/ Sanctuary AI]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 170 cm | weight = 70 kg | two_hand_payload = 25 }} [[Category:Robots]] dcd3c82313aecda5134d1f9e3d2b27e54783dbd5 243 242 2024-04-24T10:11:29Z User2024 6 wikitext text/x-wiki Phoenix is a humanoid robot from [https://sanctuary.ai/ Sanctuary AI]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 9n (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video = https://youtube.com/watch?v=FH3zbUSMAAU }} [[Category:Robots]] be293dfe2e105c119c1bb10dab2af8e0c4a3ad57 244 243 2024-04-24T10:12:52Z User2024 6 wikitext text/x-wiki Phoenix is a humanoid robot from [https://sanctuary.ai/ Sanctuary AI]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 9n (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} [[Category:Robots]] b928963d323b97d5bddd063629cc3e04cc134186 247 244 2024-04-24T15:23:38Z User2024 6 wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 9n (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} [[Category:Robots]] d7fb60f1f0f3f3734e1afcdcab51747a7bf4343b 252 247 2024-04-24T15:50:21Z 185.187.168.151 0 wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} [[Category:Robots]] f9bb8e1027b5610ce1df912a7dbc12aabf463221 0 54 245 2024-04-24T15:17:48Z User2024 6 Created page with "EVE is a humanoid robot from [[1X]]. {{infobox robot | name = EVE | organization = [[1X]] | height = 186 cm | weight = 86 kg | video_link = https://www.youtube.com/watch?v=20..." wikitext text/x-wiki EVE is a humanoid robot from [[1X]]. {{infobox robot | name = EVE | organization = [[1X]] | height = 186 cm | weight = 86 kg | video_link = https://www.youtube.com/watch?v=20GHG-R9eFI }} [[Category:Robots]] 811f8d79ffd2070e98821cb7539930a08ccffe60 0 55 246 2024-04-24T15:22:18Z User2024 6 Created page with "NEO is a humanoid robot from [[1X]]. {{infobox robot | name = NEO | organization = [[1X]] | height = 165 cm | weight = 30 kg }} [[Category:Robots]]" wikitext text/x-wiki NEO is a humanoid robot from [[1X]]. {{infobox robot | name = NEO | organization = [[1X]] | height = 165 cm | weight = 30 kg }} [[Category:Robots]] 3e01cd8490a07d14784346d9542e6eec15344f0f 0 56 248 2024-04-24T15:27:09Z User2024 6 Created page with "Sanctuary AI is building a humanoid robot called [[Phoenix]]. {{infobox company | name = Sanctuary AI | country = Canada | website_link = https://sanctuary.ai/ | robots = P..." wikitext text/x-wiki Sanctuary AI is building a humanoid robot called [[Phoenix]]. {{infobox company | name = Sanctuary AI | country = Canada | website_link = https://sanctuary.ai/ | robots = [[Phoenix]] }} [[Category:Companies]] 1de4bb8ad224b818e34d223a793b1b341413af88 0 57 249 2024-04-24T15:30:39Z User2024 6 Created page with "GR-1 is a humanoid robot from [[Fourier Intelligence]]. {{infobox robot | name = GR-1 | organization = [[1X]] | height = 165 cm | weight = 55 kg | cost = USD 149,999 }} ..." wikitext text/x-wiki GR-1 is a humanoid robot from [[Fourier Intelligence]]. {{infobox robot | name = GR-1 | organization = [[1X]] | height = 165 cm | weight = 55 kg | cost = USD 149,999 }} [[Category:Robots]] 16292cb2412b0a3148ff10ef51a6d19d398be087 250 249 2024-04-24T15:31:03Z User2024 6 wikitext text/x-wiki GR-1 is a humanoid robot from [[Fourier Intelligence]]. {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | height = 165 cm | weight = 55 kg | cost = USD 149,999 }} [[Category:Robots]] dc14560d93557db867753f3361f1471b72613f32 0 58 251 2024-04-24T15:31:36Z User2024 6 Created page with "Fourier Intelligence is building a humanoid robot called [[GR-1]]. {{infobox company | name = Fourier Intelligence | country = China | website_link = https://robots.fourierin..." wikitext text/x-wiki Fourier Intelligence is building a humanoid robot called [[GR-1]]. {{infobox company | name = Fourier Intelligence | country = China | website_link = https://robots.fourierintelligence.com/ | robots = [[GR-1]] }} [[Category:Companies]] 313af22cfa6cfa59c382767f170034b31b7fbfab 0 24 253 212 2024-04-24T15:51:30Z 185.187.168.151 0 wikitext text/x-wiki [[File:Cassie.jpg|right|200px|thumb]] Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility]]. {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = USD 250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} ==Development== On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. [[Category: Robots]] 630f71425d577b499192d0f42c5108b74fa72597 0 3 254 213 2024-04-24T15:51:41Z 185.187.168.151 0 wikitext text/x-wiki It is available for purchase [https://shop.unitree.com/products/unitree-h1 here]. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | purchase_link = https://shop.unitree.com/products/unitree-h1 }} [[Category: Robots]] 7d927bc59601d9914890744e80c136cdf5675451 0 1 255 238 2024-04-24T15:53:27Z 185.187.168.151 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 6c6e462d5b14cc5f7afe2b1d21fe29df3bea59d6 259 255 2024-04-24T18:09:50Z 136.62.52.52 0 /* Getting Started */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |} 61f03846868b7c0dad8249eaaca3f5188123251a 264 259 2024-04-24T22:21:30Z Modeless 7 Add Booster Robotics wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Era Robotics]] | [[XBot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |} 29a34f03af64a1a6997c4e3a499e3eb70af19ef3 267 264 2024-04-24T22:31:53Z Modeless 7 Era Robotics -> Robotera wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]] [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |} 80a89f569412da0552f58b9f4a896ce52fb62254 268 267 2024-04-24T22:32:13Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |} 851b9311a06e5aaedc5d39878ea59060ca8a2e58 269 268 2024-04-24T22:33:39Z Modeless 7 Agibot wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |} 69a263b2f1db30e8f5ec98f7e5106cd1d7e5b31a 274 269 2024-04-24T22:37:07Z Modeless 7 LimX Dynamics CL-1 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |} a0a57979d475a7941a70b01a25966648622c7528 276 274 2024-04-24T22:40:44Z Modeless 7 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |} 7c0b16d600c19b25098996214469dc1831edd5a0 278 276 2024-04-24T22:45:03Z AnthroDev 8 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |} 9a091e9e5fc45585f8840bbb0abf74aa2c5d5414 279 278 2024-04-24T22:46:15Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |} 33f7756b34024f686013655dc484699efaec7c1f 301 279 2024-04-24T23:56:55Z 141.156.214.27 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Proxy]] | |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |} 26d342c3471d324688849993cc926ca10d2c8fcf 302 301 2024-04-24T23:57:44Z 141.156.214.27 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Proxy]] | |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |} ef8ca9748c48057fc6de39efa44aa739ca5bc338 0 59 256 2024-04-24T15:57:34Z 185.187.168.151 0 Created page with "=== Adding an Infobox === There are a couple infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] In both, all of the fields are opti..." wikitext text/x-wiki === Adding an Infobox === There are a couple infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] In both, all of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> c6d224332ee1a97b5ce52686379d201eb6d48413 257 256 2024-04-24T15:57:45Z 185.187.168.151 0 wikitext text/x-wiki === Adding an Infobox === There are a couple infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] In both, all of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> 1d88590bc979105f3b3eaa2af51e40944613ec95 0 2 258 211 2024-04-24T16:26:50Z 185.187.168.151 0 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. See the [[Stompy Build Guide|build guide]] for a walk-through of how to build one yourself. = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] d3066bec124eb159c607c71d55cdfa0aaa3076d4 14 60 260 2024-04-24T18:19:39Z 136.62.52.52 0 Created page with "= Teleop = == whole-body == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot...." wikitext text/x-wiki = Teleop = == whole-body == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. - https://mobile-aloha.github.io/ - https://www.youtube.com/watch?v=PFw5hwNVhbA == VR teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via something akin to gRPC. - https://github.com/Improbable-AI/VisionProTeleop - https://github.com/fazildgr8/VR_communication_mujoco200 - https://github.com/pollen-robotics/reachy2021-unity-package - https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 == Tools == - Unity - gRPC - WebRTC - [[Category: Teleop]] f9a7b51373b6814e5187c1ac27603cf88d6a5f1b 261 260 2024-04-24T18:32:42Z 136.62.52.52 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). == Latency == == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC or WebRTC. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] cd133b73fc7e23a975065d60670fc0c3917ef11a 262 261 2024-04-24T18:37:27Z 136.62.52.52 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC or WebRTC. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 224ad1fc0bfb4af1a05bd89817dca10dea6b99b8 297 262 2024-04-24T23:55:02Z 141.156.214.27 0 /* VR Teleop */ wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC or WebRTC. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 5ce453d38dd38d5e509cbe37e62219bc461ef0b4 298 297 2024-04-24T23:55:16Z 141.156.214.27 0 /* Controller Teleop */ wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC or WebRTC. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 21925b264054636603453c256de0cb2c794e0b12 299 298 2024-04-24T23:55:31Z 141.156.214.27 0 /* Whole Body Teleop */ wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC or WebRTC. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 704c9bbfe08816f9e961fec353379ee80e4ca303 300 299 2024-04-24T23:56:13Z 141.156.214.27 0 /* Teleop */ wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC or WebRTC. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 1be905c192c941fd03d2b29384ce17def87ae8a2 0 61 263 2024-04-24T21:39:56Z Ben 2 Created page with "This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atop..." wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] bd3059f6d3f0352883e1768ffff230f1990fbb61 0 62 265 2024-04-24T22:23:22Z Modeless 7 Created page with "The only online presence of Booster Robotics is their YouTube channel with this video: https://www.youtube.com/watch?v=SIK2MFIIIXw" wikitext text/x-wiki The only online presence of Booster Robotics is their YouTube channel with this video: https://www.youtube.com/watch?v=SIK2MFIIIXw 7799473ae99332bc55404443ec91ed662eadc8d8 281 265 2024-04-24T22:50:03Z Modeless 7 wikitext text/x-wiki {{infobox company | name = Booster Robotics | website_link = https://www.youtube.com/watch?v=SIK2MFIIIXw | robots = [[BR002]] }} The only online presence of Booster Robotics is their YouTube channel with this video: https://www.youtube.com/watch?v=SIK2MFIIIXw 90eb4aaed57594e67fe652cde28ccda528cb4caa 282 281 2024-04-24T22:50:45Z Modeless 7 wikitext text/x-wiki The only online presence of Booster Robotics is their YouTube channel with this video: https://www.youtube.com/watch?v=SIK2MFIIIXw {{infobox company | name = Booster Robotics | website_link = https://www.youtube.com/watch?v=SIK2MFIIIXw | robots = [[BR002]] }} [[Category:Companies]] 4d02a5c4bcdb1e4e48926ade7a34444de37bcce6 1 63 266 2024-04-24T22:26:50Z Modeless 7 Created page with "You should install a YouTube plugin so we can embed robot demo videos! --~~~~" wikitext text/x-wiki You should install a YouTube plugin so we can embed robot demo videos! --[[User:Modeless|Modeless]] ([[User talk:Modeless|talk]]) 22:26, 24 April 2024 (UTC) bb3fbc881bc9f53c1da250cc8f16fd1c465d21dd 0 32 270 240 2024-04-24T22:34:26Z 104.7.66.79 0 /* Training algorithms */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks ===Isaac Gym=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===Gymnasium=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 6cb8b923d63a866774f72955ecde6ec64a97cde1 277 270 2024-04-24T22:42:33Z 104.7.66.79 0 /* Training frameworks */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 62732a33b16beafae8589a4752c2dc52ebde5f7d 291 277 2024-04-24T23:07:37Z 104.7.66.79 0 /* Simulators */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ==[[https://github.com/haosulab/ManiSkill ManiSkill]]== ===[[VSim]]=== ==Training frameworks== Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] fc31af82c9aaf95de5da92c905ac9d26bb36274b 292 291 2024-04-24T23:07:49Z 104.7.66.79 0 /* https://github.com/haosulab/ManiSkill ManiSkill */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] f53fca0a85afac92a15901c08133f171ebce011e 294 292 2024-04-24T23:11:40Z 104.7.66.79 0 /* Learning algorithms */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with possible example of applications. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 8ca58594e364c6b3fd5309125a3d5bf65bc15f2b 295 294 2024-04-24T23:12:19Z 104.7.66.79 0 /* Learning algorithms */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with example [[applications]]. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Simulators== ===[[Isaac Sim]]=== ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 9aacff7ab539b9555c4e3e55eda4670b9cfb7d6f 0 64 271 2024-04-24T22:35:45Z Modeless 7 Created page with "Agibot is a Chinese startup formed in 2022. They have a humanoid called RAISE-A1. Their CEO gave a detailed presentation about it here: https://www.youtube.com/watch?v=ZwjxbDV..." wikitext text/x-wiki Agibot is a Chinese startup formed in 2022. They have a humanoid called RAISE-A1. Their CEO gave a detailed presentation about it here: https://www.youtube.com/watch?v=ZwjxbDVbGpU&t=1471s 30e6719035c93a254d3bf112eb0f0343520274d8 280 271 2024-04-24T22:48:42Z Modeless 7 wikitext text/x-wiki Agibot is a Chinese startup formed in 2022. The name stands for "Artificial General Intelligence Bot". They have a humanoid called RAISE-A1. Their CEO gave a detailed presentation about it here: https://www.youtube.com/watch?v=ZwjxbDVbGpU&t=1471s b896c8fa1667649b273aeecba731b9bb0c7f5fde 284 280 2024-04-24T22:52:34Z Modeless 7 wikitext text/x-wiki The "Agi" in their name stands for "Artificial General Intelligence". Their CEO gave a detailed presentation here: https://www.youtube.com/watch?v=ZwjxbDVbGpU&t=1471s {{infobox company | name = Agibot | website_link = https://www.agibot.com/ | robots = [[RAISE-A1]] }} [[Category:Companies]] 52fa639517a52b0564c2195b853ebd7ca21739fb 285 284 2024-04-24T22:53:10Z Modeless 7 wikitext text/x-wiki The "Agi" in their name stands for "Artificial General Intelligence". Their CEO gave a detailed presentation here: https://www.youtube.com/watch?v=ZwjxbDVbGpU&t=1471s {{infobox company | name = Agibot | country = China | website_link = https://www.agibot.com/ | robots = [[RAISE-A1]] }} [[Category:Companies]] 73984c3a371c72de33880735a346e5eb1c3be56b 0 34 272 131 2024-04-24T22:36:25Z 104.7.66.79 0 /* PPO */ wikitext text/x-wiki ==Training algorithms== ===A2C=== ===[https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO]=== e18348cdc5f5f670cfeff48fab551f021f7b3423 273 272 2024-04-24T22:36:51Z 104.7.66.79 0 /* A2C */ wikitext text/x-wiki ==Training algorithms== ===[https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C]=== ===[https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO]=== b2a6996a393c689b8ed0d8011333bd8d9c5e9f00 0 65 275 2024-04-24T22:37:59Z Modeless 7 Created page with "LimX Dynamics was founded in 2022 in China. Their first product is a dog robot with wheels for feet, but their second product is a humanoid called "CL-1". Here is a video show..." wikitext text/x-wiki LimX Dynamics was founded in 2022 in China. Their first product is a dog robot with wheels for feet, but their second product is a humanoid called "CL-1". Here is a video showcasing their teleoperation setup to collect training data. https://www.youtube.com/watch?v=2dmjzMv-y-M ece378ea8022c228973ffb1bc02d3a5c0121654f 286 275 2024-04-24T22:54:41Z Modeless 7 wikitext text/x-wiki LimX Dynamics was founded in 2022 in China. Their first product is a dog robot with wheels for feet, but their second product is a humanoid called [[CL-1]]. Here is a video showcasing their teleoperation setup to collect training data. https://www.youtube.com/watch?v=2dmjzMv-y-M {{infobox company | name = LimX Dynamics | country = China | website_link = https://www.limxdynamics.com/en | robots = [[CL-1]] }} [[Category:Companies]] d2703b0dfbcda8e0d84361b129edefcacbd2aaf4 0 66 283 2024-04-24T22:51:25Z 104.7.66.79 0 Created page with "=Applications List= A non-comprehensive list of training frameworks is listed below. ===[https://github.com/leggedrobotics/legged_gym Legged Gym]=== Isaac Gym Environments fo..." wikitext text/x-wiki =Applications List= A non-comprehensive list of training frameworks is listed below. ===[https://github.com/leggedrobotics/legged_gym Legged Gym]=== Isaac Gym Environments for Legged Robots. ===[https://github.com/roboterax/humanoid-gym Humanoid Gym]=== Training setup for walking with [[Xbot-L]]. ===[https://github.com/kscalelabs/sim KScale Sim]=== Training setup for getting up and walking with [[Stompy]]. dcc27dbe09127c2a7c87798dfca8c958362d4b50 287 283 2024-04-24T22:57:36Z 104.7.66.79 0 /* Applications List */ wikitext text/x-wiki =Applications List= A non-comprehensive list of training frameworks is listed below. ===[https://github.com/leggedrobotics/legged_gym Legged Gym]=== Isaac Gym Environments for Legged Robots. ===[https://github.com/Alescontrela/AMP_for_hardware AMP for hardware]=== Adversarial Motion Priors Make Good Substitutes for Complex Reward Functions ==[https://github.com/ZhengyiLuo/PHC PHC]=== Official Implementation of the ICCV 2023 paper: Perpetual Humanoid Control for Real-time Simulated Avatars. ===[https://github.com/roboterax/humanoid-gym Humanoid Gym]=== Training setup for walking with [[Xbot-L]]. ===[https://github.com/kscalelabs/sim KScale Sim]=== Training setup for getting up and walking with [[Stompy]]. 0af6e8829cc616c85c3bad05d9939c1d586b70b1 288 287 2024-04-24T22:58:00Z 104.7.66.79 0 wikitext text/x-wiki =Applications List= A non-comprehensive list of training frameworks is listed below. ===[https://github.com/leggedrobotics/legged_gym Legged Gym]=== Isaac Gym Environments for Legged Robots. ===[https://github.com/Alescontrela/AMP_for_hardware AMP for hardware]=== Adversarial Motion Priors Make Good Substitutes for Complex Reward Functions ===[https://github.com/ZhengyiLuo/PHC PHC]=== Official Implementation of the ICCV 2023 paper: Perpetual Humanoid Control for Real-time Simulated Avatars. ===[https://github.com/roboterax/humanoid-gym Humanoid Gym]=== Training setup for walking with [[Xbot-L]]. ===[https://github.com/kscalelabs/sim KScale Sim]=== Training setup for getting up and walking with [[Stompy]]. 78eb6b45fa1c3aee693dcebe08f52972e66a293f 289 288 2024-04-24T22:59:15Z 104.7.66.79 0 wikitext text/x-wiki =Applications List= A non-comprehensive list of training frameworks is listed below. ===[https://github.com/leggedrobotics/legged_gym Legged Gym]=== Isaac Gym Environments for Legged Robots. ===[https://github.com/chengxuxin/extreme-parkour Extreme Parkour]=== Extreme Parkour with AMP Legged Robots. ===[https://github.com/Alescontrela/AMP_for_hardware AMP for hardware]=== Adversarial Motion Priors Make Good Substitutes for Complex Reward Functions ===[https://github.com/ZhengyiLuo/PHC PHC]=== Official Implementation of the ICCV 2023 paper: Perpetual Humanoid Control for Real-time Simulated Avatars. ===[https://github.com/roboterax/humanoid-gym Humanoid Gym]=== Training setup for walking with [[Xbot-L]]. ===[https://github.com/kscalelabs/sim KScale Sim]=== Training setup for getting up and walking with [[Stompy]]. 0ed84ddf4368f699b5074eb72406cf103c4d9f10 6 67 290 2024-04-24T23:01:35Z AnthroDev 8 Logo for Anthrobotics wikitext text/x-wiki == Summary == Logo for Anthrobotics cf09f4204755c59510c0154435653be09feb1240 0 68 293 2024-04-24T23:10:01Z AnthroDev 8 Created page with "[[File:Logo Anthrobotics Full.jpeg|right|400px|thumb]] Anthrobotics is an Alberta-based startup company developing open-source humanoid robots and deployable autonomous syste..." wikitext text/x-wiki [[File:Logo Anthrobotics Full.jpeg|right|400px|thumb]] Anthrobotics is an Alberta-based startup company developing open-source humanoid robots and deployable autonomous systems. Their flagship project in development is a digitigrade humanoid robot called the [[Anthro]] (short for Anthropomorphic Robot). {{infobox company | name = Anthrobotics | country = Canada | website_link = https://anthrobotics.ca/ | robots = [[Anthro]] }} Anthrobotics was founded in 2016 as a continuation of "Project '87", a 2014 collaborative community project centered around building real-life walking replicas of the animatronic characters from the Five Nights at Freddy's video game series. As the animatronics required a full-size humanoid robot to perform as mobile units rather than static displays, the scope of the project was expanded to develop a general purpose, highly customizable humanoid. Development of the Anthro is ongoing, with the end goal of providing users and developers with a modular, highly customizable humanoid that can take on a diverse range of anthropomorphic appearances. The Anthro is intended to be used in any general purpose application, or in niche roles. [[Category:Companies]] e4385b451e770c4e61398a6f04bb58cafe014f71 2 69 296 2024-04-24T23:21:14Z Modeless 7 Created page with "Hi I'm James Darpinian. Check my website: https://james.darpinian.com/ Follow me on X: https://x.com/modeless" wikitext text/x-wiki Hi I'm James Darpinian. Check my website: https://james.darpinian.com/ Follow me on X: https://x.com/modeless 98560545d53c8d0d13ce62da2291aa7a74680b16 0 1 303 302 2024-04-24T23:59:02Z 141.156.214.27 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] | |- | [[Proxy]] |} 067611ad6c239f6d7e5440d0f493d3d9c3f0a048 304 303 2024-04-24T23:59:55Z 141.156.214.27 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] | |- | [[Proxy]] | |} b45fb8e5014ea9e9017a9e3959aa1963b7921ab7 343 304 2024-04-25T19:41:14Z 24.130.242.94 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[Proxy]] | |} f07457b16e2bdb603295df117896c45ce7e5aeb7 346 343 2024-04-25T21:14:23Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[Proxy]] | |} 97be954cc43be1b374add9b69f5021544cc4a81b 0 70 305 2024-04-25T00:04:25Z 141.156.214.27 0 Created page with "Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. Unlike autonomous robots that rely on pre-programmed instructions or arti..." wikitext text/x-wiki Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, several existing robotics firms are partnering with Proxy developers to incorporate this human-centric approach into their humanoid robots. To learn more about Proxy technology and potential collaborations, contact contact@what-is-proxy.com. f22ab625a74201c818b38a983d567e81f036b394 306 305 2024-04-25T00:07:00Z 141.156.214.27 0 Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. wikitext text/x-wiki == Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. == Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, existing robotics firms are partnering with Proxy to incorporate their human-centric approach into their humanoid robots. To learn more about Proxy technology and inquire about potential collaborations, their website provides contact@what-is-proxy.com. c43f4b57334c2328dd8b50412c44dfa67eeaf853 307 306 2024-04-25T00:08:19Z 141.156.214.27 0 wikitext text/x-wiki == Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. == Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, existing robotics firms are partnering with Proxy to incorporate their human-centric approach into their humanoid robots. To learn more about Proxy technology and inquire about potential collaborations, email contact@what-is-proxy.com. 2c6fbccb92c50f8cfcc2773c6c0a5daed8a304f4 308 307 2024-04-25T00:14:23Z 141.156.214.27 0 /* Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. */ wikitext text/x-wiki == Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. == Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, existing robotics firms are partnering with Proxy to incorporate their human-centric approach into their humanoid robots. Comments from the team: * To learn more about Proxy technology, visit https://https://what-is-proxy.com * To inquire about potential collaborations, email '''contact@what-is-proxy.com'''. 18e4e22a480d31d9e8f9ca680f025a391bae2318 309 308 2024-04-25T00:24:47Z 141.156.214.27 0 wikitext text/x-wiki == Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. == Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, existing robotics firms are partnering with Proxy to incorporate their human-centric approach into their humanoid robots. Comments from the team: * To learn more about Proxy, visit https://https://what-is-proxy.com 6c66be3a1327ee40c3b410edac3b2b0f84e58390 310 309 2024-04-25T00:25:36Z 141.156.214.27 0 wikitext text/x-wiki == Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. == Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, existing robotics firms are partnering with Proxy to incorporate their human-centric approach into their humanoid robots. Comments from the team: * To learn more about Proxy, visit https://what-is-proxy.com 385171bae11ba5b6ad65d13b9025979a1e028bad 338 310 2024-04-25T17:28:53Z Mrroboto 5 wikitext text/x-wiki == Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. == Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, existing robotics firms are partnering with Proxy to incorporate their human-centric approach into their humanoid robots. Comments from the team: * To learn more about Proxy, visit https://what-is-proxy.com [[Category: Teleop]] 5cdfc77741d3dcb88878963ae49980442c3f45bc 339 338 2024-04-25T17:29:17Z Mrroboto 5 wikitext text/x-wiki === About === Proxy refers to a category of robots specifically designed for real-time, remote operation by humans. === Description === Unlike autonomous robots that rely on pre-programmed instructions or artificial intelligence for decision-making, Proxy robots act as physical extensions of their human operators, mimicking their movements and actions with minimal latency. This technology allows humans to experience and interact with the physical world from afar, opening up new possibilities for remote work, exploration, and accessibility. Proxy robots utilize various methods for control, including virtual reality headsets and haptic feedback systems, providing operators with an immersive and intuitive experience. While Proxy technology is still in its early stages, existing robotics firms are partnering with Proxy to incorporate their human-centric approach into their humanoid robots. Comments from the team: * To learn more about Proxy, visit https://what-is-proxy.com [[Category: Teleop]] 6c987a4ffbd5a3d39a40059ab1c95bb0018318c7 0 71 311 2024-04-25T03:58:32Z User2024 6 Created page with "Walker X is a humanoid robot from [[UBTech]]. {{infobox robot | name = Walker X | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1.5 | t..." wikitext text/x-wiki Walker X is a humanoid robot from [[UBTech]]. {{infobox robot | name = Walker X | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1.5 | two_hand_payload = 3 | cost = USD 960,000 }} [[Category:Robots]] cec319bba5b1aaf629eb63538c59ab9723e34768 313 311 2024-04-25T04:00:52Z User2024 6 wikitext text/x-wiki Walker X is a humanoid robot from [[UBTech]]. {{infobox robot | name = Walker X | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1.5 | two_hand_payload = 3 | cost = USD 960,000 | video_link = https://www.youtube.com/watch?v=4ZL3LgdKNbw }} [[Category:Robots]] 22663b29dd7050cade911ac50b93b6913669f849 0 72 312 2024-04-25T04:00:17Z User2024 6 Created page with "UBTech is building a humanoid robot called [[Walker X]]. {{infobox company | name = UBTech | country = China | website_link = https://www.ubtrobot.com/ | robots = [[Walker X]..." wikitext text/x-wiki UBTech is building a humanoid robot called [[Walker X]]. {{infobox company | name = UBTech | country = China | website_link = https://www.ubtrobot.com/ | robots = [[Walker X]] }} [[Category:Companies]] 10f13a655e2e68a1b8f2dc0baf64230aacc592b6 315 312 2024-04-25T04:05:28Z User2024 6 wikitext text/x-wiki UBTech is building a humanoid robot called [[Walker X]], [[Panda Robot]]. {{infobox company | name = UBTech | country = China | website_link = https://www.ubtrobot.com/ | robots = [[Walker X]], [[Panda Robot]] }} [[Category:Companies]] 6d9565187459529a208f31d5732d5782d50b446c 316 315 2024-04-25T04:06:04Z User2024 6 wikitext text/x-wiki UBTech is building humanoid robots called [[Walker X]], [[Panda Robot]]. {{infobox company | name = UBTech | country = China | website_link = https://www.ubtrobot.com/ | robots = [[Walker X]], [[Panda Robot]] }} [[Category:Companies]] 249e261ca0940994da7ae03206bad04c5e052512 318 316 2024-04-25T04:07:53Z User2024 6 wikitext text/x-wiki UBTech is building humanoid robots called [[Walker X]], [[Panda Robot]] and [[Walker S]]. {{infobox company | name = UBTech | country = China | website_link = https://www.ubtrobot.com/ | robots = [[Walker X]], [[Panda Robot]], [[Walker S]] }} [[Category:Companies]] 8728a8c97c3059a387e5997f681bc78192c44dbe 0 73 314 2024-04-25T04:04:10Z User2024 6 Created page with "Panda Robot is a humanoid robot from [[UBTech]]. {{infobox robot | name = Panda Robot | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1..." wikitext text/x-wiki Panda Robot is a humanoid robot from [[UBTech]]. {{infobox robot | name = Panda Robot | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1.5 | two_hand_payload = 3 | cost = USD 960,000 }} [[Category:Robots]] 381f47cd90976c0fb6c30f49d2d8622695225a90 0 74 317 2024-04-25T04:07:23Z User2024 6 Created page with "Walker S is a humanoid robot from [[UBTech]]. {{infobox robot | name = Walker S | organization = [[UBTech]] | height = | weight = | single_hand_payload = | two_hand_payload..." wikitext text/x-wiki Walker S is a humanoid robot from [[UBTech]]. {{infobox robot | name = Walker S | organization = [[UBTech]] | height = | weight = | single_hand_payload = | two_hand_payload = | cost = }} [[Category:Robots]] 4155935e48cc89cc22aa75ae2be305d68fa0a504 0 75 319 2024-04-25T05:11:35Z User2024 6 Created page with "Wukong-IV is a humanoid robot from [[Deep Robotics]]. {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_p..." wikitext text/x-wiki Wukong-IV is a humanoid robot from [[Deep Robotics]]. {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_payload | two_hand_payload | cost = | video_link = https://www.youtube.com/watch?v=fbk4fYc6U14 }} [[Category:Robots]] 2f89a47f8314e7f8d95d970a117bdbdfbf105bcc 0 76 320 2024-04-25T05:19:17Z User2024 6 Created page with "Deep Robotics is building a humanoid robot called [[Wukong-IV]]. {{infobox company | name = Deep Robotics | country = China | website_link = https://www.deeprobotics.cn/en |..." wikitext text/x-wiki Deep Robotics is building a humanoid robot called [[Wukong-IV]]. {{infobox company | name = Deep Robotics | country = China | website_link = https://www.deeprobotics.cn/en | robots = [[Wukong-IV]] }} [[Category:Companies]] 4e74e6b2665bebe1006fcf7077e22f4c2be72a9c 0 77 321 2024-04-25T05:44:52Z User2024 6 Created page with "XR4 is a humanoid robot from [[DATAA Robotics]]. {{infobox robot | name = XR4 | organization = [[DATAA Robotics]] | height = | weight = | single_hand_payload | two_hand_pa..." wikitext text/x-wiki XR4 is a humanoid robot from [[DATAA Robotics]]. {{infobox robot | name = XR4 | organization = [[DATAA Robotics]] | height = | weight = | single_hand_payload | two_hand_payload | cost = }} [[Category:Robots]] 3213a6fde6a5f433249dbb03515161ff6f1d3704 0 78 322 2024-04-25T05:45:15Z User2024 6 Created page with "DATAA Robotics is building a humanoid robot called [[XR4]]. {{infobox company | name = DATAA Robotics | country = China | website_link = https://www.dataarobotics.com/en | ro..." wikitext text/x-wiki DATAA Robotics is building a humanoid robot called [[XR4]]. {{infobox company | name = DATAA Robotics | country = China | website_link = https://www.dataarobotics.com/en | robots = [[XR4]] }} [[Category:Companies]] a416f7e2f439bfa56326e6fe290e0377d3b20502 0 79 323 2024-04-25T05:56:25Z User2024 6 Created page with "ZEUS2Q is a humanoid robot from [[System Technology Works]]. {{infobox robot | name = ZEUS2Q | organization = [[System Technology Works]] | height = 127 cm | weight = 13.61 k..." wikitext text/x-wiki ZEUS2Q is a humanoid robot from [[System Technology Works]]. {{infobox robot | name = ZEUS2Q | organization = [[System Technology Works]] | height = 127 cm | weight = 13.61 kg }} [[Category:Robots]] 9e448ace10e1576a040a0c29823e6e2e9493150b 0 80 324 2024-04-25T05:56:50Z User2024 6 Created page with "System Technology Works is building a humanoid robot called [[ZEUS2Q]]. {{infobox company | name = System Technology Works | country = USA | website_link = https://www.system..." wikitext text/x-wiki System Technology Works is building a humanoid robot called [[ZEUS2Q]]. {{infobox company | name = System Technology Works | country = USA | website_link = https://www.systemtechnologyworks.com/ | robots = [[ZEUS2Q]] }} [[Category:Companies]] 8184bac9f7f1c1c640ceaabf86451936ec281279 0 18 325 56 2024-04-25T07:16:20Z Stonet2000 9 wikitext text/x-wiki Isaac Sim is a simulator from NVIDIA connect with the Omniverse platform. The core physics engine underlying Isaac Sim is PhysX === Doing Simple Operations === '''Start Isaac Sim''' * Open Omniverse Launcher * Navigate to the Library * Under “Apps” click “Isaac Sim” * Click “Launch” * There are multiple options for launching. Choose the normal one to show the GUI or headless if streaming. * Choose <code>File > Open...</code> and select the <code>.usd</code> model corresponding to the robot you want to simulate. '''Connecting streaming client''' * Start Isaac Sim in Headless (Native) mode * Open Omniverse Streaming Client * Connect to the server [[Category: Simulators]] e1f470c78f865948ede2d6e32caa9a81526abfa4 342 325 2024-04-25T19:40:42Z 24.130.242.94 0 wikitext text/x-wiki Isaac Sim is a simulator from NVIDIA connect with the Omniverse platform. The core physics engine underlying Isaac Sim is PhysX === Doing Simple Operations === '''Start Isaac Sim''' * Open Omniverse Launcher * Navigate to the Library * Under “Apps” click “Isaac Sim” * Click “Launch” ** There are multiple options for launching. Choose the normal one to show the GUI or headless if streaming. * Choose <code>File > Open...</code> and select the <code>.usd</code> model corresponding to the robot you want to simulate. '''Connecting streaming client''' * Start Isaac Sim in Headless (Native) mode * Open Omniverse Streaming Client * Connect to the server [[Category: Simulators]] c34d38192832cd8dbf7eb2b98a99deb9c5eae494 0 32 326 295 2024-04-25T07:23:39Z Stonet2000 9 /* Physics engines */ wikitext text/x-wiki = Learning algorithms = Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with example [[applications]]. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ==Bullet== Bullet is a physics engine supporting real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning ==Simulators== ===[[Isaac Sim]]=== ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== ==Training frameworks== Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 56e8987d2438cc98a87de485f22761d959498d57 341 326 2024-04-25T19:30:09Z 206.0.71.44 0 /* Learning algorithms */ wikitext text/x-wiki Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with example [[applications]]. Typically you need a simulator, training framework and machine learning method to train end to end behaviors. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. For a much more comprehensive overview see [https://simulately.wiki/docs/ Simulately]. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ===Bullet=== Bullet is a physics engine supporting real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning ==Simulators== ===[[Isaac Sim]]=== ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== =Training frameworks= Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 2860a19ac7cf857bd829ac7d1b945c47d4afeb94 0 81 327 2024-04-25T15:18:45Z User2024 6 Created page with "Atlas is a humanoid robot from [[Boston Dynamics]]. {{infobox robot | name = Atlas | organization = [[System Technology Works]] | height = 150 cm | weight = 89 kg | single_ha..." wikitext text/x-wiki Atlas is a humanoid robot from [[Boston Dynamics]]. {{infobox robot | name = Atlas | organization = [[System Technology Works]] | height = 150 cm | weight = 89 kg | single_hand_payload | two_hand_payload | cost = }} [[Category:Robots]] 5f953a3de0ea59876ba24e9d7912c57d5cce13c2 328 327 2024-04-25T15:19:20Z User2024 6 wikitext text/x-wiki Atlas is a humanoid robot from [[Boston Dynamics]]. {{infobox robot | name = Atlas | organization = [[Boston Dynamics]] | height = 150 cm | weight = 89 kg | single_hand_payload | two_hand_payload | cost = }} [[Category:Robots]] 2cc9b28fef32d25c31b8b92302082ad76f09f3bb 330 328 2024-04-25T15:20:47Z User2024 6 wikitext text/x-wiki Atlas is a humanoid robot from [[Boston Dynamics]]. {{infobox robot | name = Atlas | organization = [[Boston Dynamics]] | height = 150 cm | weight = 89 kg | video_link = https://www.youtube.com/watch?v=29ECwExc-_M | single_hand_payload | two_hand_payload | cost = }} [[Category:Robots]] 5e5d99ddbecaa0d4c8e7bb85349d9ca9e1da44e5 0 82 329 2024-04-25T15:19:36Z User2024 6 Created page with "Boston Dynamics is building a humanoid robot called [[Atlas]]. {{infobox company | name = Boston Dynamics | country = USA | website_link = https://bostondynamics.com/ | robot..." wikitext text/x-wiki Boston Dynamics is building a humanoid robot called [[Atlas]]. {{infobox company | name = Boston Dynamics | country = USA | website_link = https://bostondynamics.com/ | robots = [[Atlas]] }} [[Category:Companies]] 5c887274a6346f40f073bfcb90fc8cdbaec1a4a9 0 83 331 2024-04-25T15:24:53Z User2024 6 Created page with "HUBO is a humanoid robot from [[Rainbow Robotics]]. {{infobox robot | name = HUBO | organization = [[Rainbow Robotics]] | height = 170 cm | weight = 80 kg | single_hand_paylo..." wikitext text/x-wiki HUBO is a humanoid robot from [[Rainbow Robotics]]. {{infobox robot | name = HUBO | organization = [[Rainbow Robotics]] | height = 170 cm | weight = 80 kg | single_hand_payload | two_hand_payload | cost = USD 320,000 }} [[Category:Robots]] 69530615f7ead20f43ea83dd268609739f2e2a84 333 331 2024-04-25T15:27:07Z User2024 6 wikitext text/x-wiki HUBO is a humanoid robot from [[Rainbow Robotics]]. {{infobox robot | name = HUBO | organization = [[Rainbow Robotics]] | height = 170 cm | weight = 80 kg | video_link = https://www.youtube.com/watch?v=r2pKEVTddy4 | single_hand_payload | two_hand_payload | cost = USD 320,000 }} [[Category:Robots]] dcd976f2bf1ef677404c4a1aa7829410a5926075 0 84 332 2024-04-25T15:25:31Z User2024 6 Created page with "Rainbow Robotics is building a humanoid robot called [[HUBO]]. {{infobox company | name = Rainbow Robotics | country = South Korea | website_link = https://www.rainbow-roboti..." wikitext text/x-wiki Rainbow Robotics is building a humanoid robot called [[HUBO]]. {{infobox company | name = Rainbow Robotics | country = South Korea | website_link = https://www.rainbow-robotics.com/ | robots = [[HUBO]] }} [[Category:Companies]] f411e775cf99891b465c825915ae1d1544488ce8 0 53 334 252 2024-04-25T17:25:30Z Ben 2 wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[Category:Robots]] d7f7c22b8bd19ab3101e0e9897b3273c0e3613c2 336 334 2024-04-25T17:27:07Z Mrroboto 5 wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[File:Main-image-phoenix-annoucement.jpg|thumb]] [[Category:Robots]] 0e39bdb32d07df4471ea468c91e0a8c378758369 337 336 2024-04-25T17:27:44Z Mrroboto 5 wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] [[Category:Robots]] 0812ff6b58d7c71aa8fff8323e9461fd5fd0707a 6 85 335 2024-04-25T17:26:59Z Mrroboto 5 wikitext text/x-wiki Phoenix Gen 7 press release image adb64e395b5464b8cf19a4215f2160c1668ad7e2 0 34 340 273 2024-04-25T18:56:07Z 2.127.48.230 0 /* Training algorithms */ wikitext text/x-wiki ==Training algorithms== ===[https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C]=== ===[https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO]=== ===[https://spinningup.openai.com/en/latest/algorithms/sac.html SAC]=== e4afe1e8be68aaa87d905516fff02e9d7724ffb2 0 65 344 286 2024-04-25T20:16:45Z 24.130.242.94 0 wikitext text/x-wiki LimX Dynamics was founded in 2022 in China. Their first product is a dog robot with wheels for feet, but their second product is a humanoid called [[CL-1]]. [https://www.youtube.com/watch?v=2dmjzMv-y-M Here is a video] showcasing their teleoperation setup to collect training data. {{infobox company | name = LimX Dynamics | country = China | website_link = https://www.limxdynamics.com/en | robots = [[CL-1]] }} [[Category:Companies]] e623a16bfe4b575e99dc192c02535fb1d8b84570 0 61 345 263 2024-04-25T21:13:27Z Ben 2 wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] === Commercial Servos === * [https://www.myactuator.com/downloads-x-series RMD X Series] ** Based on the MIT Cheetah actuator ** Planetary gears with 1:6 to 1:8 reduction ratios === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] 72aa1ee0d9a9aaef0e0f1ea34bbdd94a664ca065 347 345 2024-04-25T21:17:30Z Ben 2 wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] === Commercial Servos === * [https://www.myactuator.com/downloads-x-series MyActuator X Series] [[MyActuator X-Series]] ** Based on the MIT Cheetah actuator ** Planetary gears with 1:6 to 1:8 reduction ratios === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] 83bcff1ca1e1244b6ea8d3593f8cefe0bd8f542b 10 86 348 2024-04-25T21:19:50Z Ben 2 Created page with "{{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Price | value3 = {{{price}}} | key4 = Purchase..." wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Price | value3 = {{{price}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Torque | value5 = {{{torque|}}} | key6 = Mass | value6 = {{{mass|}}} | key7 = CAD | value7 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} | key8 = Size | value8 = {{{size}}} }} 58bef1b36076a70d640462ef9b73f017817bde42 349 348 2024-04-25T21:20:00Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name|}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Price | value3 = {{{price|}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Torque | value5 = {{{torque|}}} | key6 = Mass | value6 = {{{mass|}}} | key7 = CAD | value7 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} | key8 = Size | value8 = {{{size|}}} }} f22d5e96a71156420a7d7e4e0d39f99971caf85f 350 349 2024-04-25T21:20:15Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Price | value3 = {{{price|}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Torque | value5 = {{{torque|}}} | key6 = Mass | value6 = {{{mass|}}} | key7 = CAD | value7 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} | key8 = Size | value8 = {{{size|}}} }} 3ef51852acd16ae46f72eed212ea6581a637f308 351 350 2024-04-25T21:22:19Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Cost | value3 = {{{cost|}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Torque | value5 = {{{torque|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = CAD | value7 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} | key8 = Size | value8 = {{{size|}}} }} 635a717a64015a08b816652e1a2a38e02da3b549 0 59 352 257 2024-04-25T21:22:23Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | torque = 1 Nm | weight = 1 kg | cad_link = https://example.com/model.step | size = 10cm radius }} </syntaxhighlight> 3f382668d5ce61a8aebf74b11b2027f33d7f2d76 0 59 353 352 2024-04-25T21:22:43Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | torque = 1 Nm | weight = 1 kg | cad_link = https://example.com/model.step | size = 10cm radius }} </syntaxhighlight> c5b93881a1150b75249306d62d8f8b2ee1b31d12 355 353 2024-04-25T21:26:45Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | nominal_torque = 1 Nm | peak_torque = 1 Nm | weight = 1 kg | size = 10cm radius | gear_ratio = 1:8 | voltage = 48V | cad_link = https://example.com/model.step }} </syntaxhighlight> 4e324da542b23672bcefa4146e2848d7f19612ec 357 355 2024-04-25T21:28:47Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | nominal_torque = 1 Nm | peak_torque = 1 Nm | weight = 1 kg | size = 10cm radius | gear_ratio = 1:8 | voltage = 48V | cad_link = https://example.com/model.step | interface = CAN }} </syntaxhighlight> 17033048c684b22dc1a0e2869d8528d2a59d6986 360 357 2024-04-25T21:30:21Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | nominal_torque = 1 Nm | peak_torque = 1 Nm | weight = 1 kg | size = 10cm radius | gear_ratio = 1:8 | voltage = 48V | cad_link = https://example.com/model.step | interface = CAN | gear_type = Planetary }} </syntaxhighlight> 4f6bda509f5b849a58c086ab410f05caa0214a22 366 360 2024-04-25T21:34:14Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <name>. To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | nominal_torque = 1 Nm | peak_torque = 1 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | cad_link = https://example.com/model.step | interface = CAN | gear_type = Planetary }} </syntaxhighlight> 54148cd7a2d02bc6896e4d834348028d76ad2ce7 382 366 2024-04-25T22:22:11Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <name>. === Company === To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox company | name = | country = | website_link = | robots = }} </syntaxhighlight> === Robot === To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox robot | name = | organization = | video_link = | cost = | height = | weight = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} </syntaxhighlight> === Actuator === To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | nominal_torque = 1 Nm | peak_torque = 1 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | cad_link = https://example.com/model.step | interface = CAN | gear_type = Planetary }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox actuator | name = | manufacturer = | cost = | purchase_link = | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = | gear_type = }} </syntaxhighlight> 878f0457124f15091d7342af0104814efcb2850a 383 382 2024-04-25T22:22:28Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <code>name</code>. === Company === To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox company | name = | country = | website_link = | robots = }} </syntaxhighlight> === Robot === To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox robot | name = | organization = | video_link = | cost = | height = | weight = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} </syntaxhighlight> === Actuator === To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | nominal_torque = 1 Nm | peak_torque = 1 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | cad_link = https://example.com/model.step | interface = CAN | gear_type = Planetary }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox actuator | name = | manufacturer = | cost = | purchase_link = | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = | gear_type = }} </syntaxhighlight> 26714788e3f6746d7335b11bec9f6a1f454afc55 384 383 2024-04-25T22:22:42Z Ben 2 wikitext text/x-wiki === Adding an Infobox === There are a few infobox templates which we use: * [[Template:Infobox_company]] * [[Template:Infobox_robot]] * [[Template:Infobox_actuator]] All of the fields are optional, except for <code>name</code>. === Company === To add a company infobox, use: <syntaxhighlight lang="text"> {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox company | name = | country = | website_link = | robots = }} </syntaxhighlight> === Robot === To add a robot infobox, use: <syntaxhighlight lang="text"> {{infobox robot | name = Name | organization = Company | video_link = https://youtube.com/ | cost = USD 1 | height = 10 ft | weight = 100 kg | speed = 1 m/s | lift_force = 10 lb | battery_life = 5 hr | battery_capacity = 1 mWh | purchase_link = https://buy.com | number_made = 10 | dof = 10 | status = Available }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox robot | name = | organization = | video_link = | cost = | height = | weight = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} </syntaxhighlight> === Actuator === To add an actuator infobox, use: <syntaxhighlight lang="text"> {{infobox actuator | name = Name | manufacturer = Manufacturer | cost = USD 1 | purchase_link = https://amazon.com/ | nominal_torque = 1 Nm | peak_torque = 1 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | cad_link = https://example.com/model.step | interface = CAN | gear_type = Planetary }} </syntaxhighlight> Here is the blank version: <syntaxhighlight lang="text"> {{infobox actuator | name = | manufacturer = | cost = | purchase_link = | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = | gear_type = }} </syntaxhighlight> e52333e859f5768dee12a579ed8d60cae29a7ffb 10 86 354 351 2024-04-25T21:25:55Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Cost | value3 = {{{cost|}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Nominal Torque | value5 = {{{nominal_torque|}}} | key6 = Peak Torque | value6 = {{{peak_torque|}}} | key7 = Weight | value7 = {{{weight|}}} | key8 = Size | value8 = {{{size|}}} | key9 = Gear Ratio | value9 = {{{gear_ratio|}}} | key10 = Voltage | value10 = {{{voltage|}}} | key11 = CAD | value11 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} }} 4bb7840ba9f7aa697fa43305d06541956ae925c3 356 354 2024-04-25T21:28:37Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Cost | value3 = {{{cost|}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Nominal Torque | value5 = {{{nominal_torque|}}} | key6 = Peak Torque | value6 = {{{peak_torque|}}} | key7 = Weight | value7 = {{{weight|}}} | key8 = Size | value8 = {{{size|}}} | key9 = Gear Ratio | value9 = {{{gear_ratio|}}} | key10 = Voltage | value10 = {{{voltage|}}} | key11 = CAD | value11 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} | key12 = Interface | value12 = {{{interface|}}} }} b95e325a200386749438caa4fb3ce94bf98e1f3e 359 356 2024-04-25T21:30:09Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Cost | value3 = {{{cost|}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Nominal Torque | value5 = {{{nominal_torque|}}} | key6 = Peak Torque | value6 = {{{peak_torque|}}} | key7 = Weight | value7 = {{{weight|}}} | key8 = Size | value8 = {{{size|}}} | key9 = Gear Ratio | value9 = {{{gear_ratio|}}} | key10 = Voltage | value10 = {{{voltage|}}} | key11 = CAD | value11 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} | key12 = Interface | value12 = {{{interface|}}} | key13 = Gear Type | value13 = {{{gear_type|}}} }} 5d2339f5b0278b09dbda9c2e16251b22534c778a 365 359 2024-04-25T21:34:06Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Manufacturer | value2 = {{{manufacturer|}}} | key3 = Cost | value3 = {{{cost|}}} | key4 = Purchase | value4 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key5 = Nominal Torque | value5 = {{{nominal_torque|}}} | key6 = Peak Torque | value6 = {{{peak_torque|}}} | key7 = Weight | value7 = {{{weight|}}} | key8 = Dimensions | value8 = {{{dimensions|}}} | key9 = Gear Ratio | value9 = {{{gear_ratio|}}} | key10 = Voltage | value10 = {{{voltage|}}} | key11 = CAD | value11 = {{#if: {{{cad_link|}}} | [{{{cad_link}}} Link] }} | key12 = Interface | value12 = {{{interface|}}} | key13 = Gear Type | value13 = {{{gear_type|}}} }} 8dae25613298696425e69eb00e94f3ea05caee15 0 61 368 347 2024-04-25T21:38:09Z Ben 2 wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] === Commercial Servos === * [[MyActuator X-Series]] ** Based on the MIT Cheetah actuator ** Planetary gears with 1:6 to 1:8 reduction ratios === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] 71575e1ec9d695a12dea2fd9928c9878906de49f 370 368 2024-04-25T21:39:36Z Ben 2 wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] === Commercial Servos === * [[MyActuator X-Series]] === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] 44c255afcbd2dcc9cdbc2869e315db16b7643d74 374 370 2024-04-25T21:44:08Z Ben 2 wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] === Commercial Servos === * [[MyActuator X-Series]] === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] [[Category: Actuators]] 5a39de92d02d7995c735b97c188d4927aad44abe 393 374 2024-04-26T04:34:01Z 24.143.251.200 0 /* Open Source Servos */ wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] * [https://github.com/jcchurch13/Mechaduino-Firmware] === Commercial Servos === * [[MyActuator X-Series]] === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] [[Category: Actuators]] d81c3e9d117824b04fa838be38f372354c366684 394 393 2024-04-26T04:34:27Z 24.143.251.200 0 /* Open Source Servos */ wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] * [https://github.com/jcchurch13/Mechaduino-Firmware Mechaduino-Firmware] === Commercial Servos === * [[MyActuator X-Series]] === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] [[Category: Hardware]] [[Category: Electronics]] [[Category: Actuators]] 7e739d24b0350647246e202e93f76439a3bde2df 0 1 369 346 2024-04-25T21:39:04Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[Proxy]] | |} 6de6811f280a8a17a373ccd3052b7191df7f8980 378 369 2024-04-25T22:04:38Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[Proxy]] | |} 359975aa537be3b7cfc245af75cd93ca69af4145 379 378 2024-04-25T22:15:27Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[Proxy]] | |} 35c8766e4ac49aebf4e133782cdf80e8101b8fa9 14 88 376 2024-04-25T21:44:39Z Ben 2 Created page with "This is a category to use for anything related to actuators." wikitext text/x-wiki This is a category to use for anything related to actuators. ce3d069b2b9b4cf8c4f515982dde5316e6b62563 0 62 377 282 2024-04-25T21:46:30Z Ben 2 wikitext text/x-wiki The only online presence of Booster Robotics is their YouTube channel with [https://www.youtube.com/watch?v=SIK2MFIIIXw this video]. {{infobox company | name = Booster Robotics | website_link = https://www.youtube.com/watch?v=SIK2MFIIIXw | robots = [[BR002]] }} [[Category:Companies]] 7e7c868bcc2414124de6453e3ad5a66a8e64ae2b 0 89 380 2024-04-25T22:18:08Z Ben 2 Created page with "The [https://github.com/unhuman-io/obot OBot] is an open-source robot designed for doing mobile manipulation. {{infobox actuator | name = OBot | voltage = 36V | interface = U..." wikitext text/x-wiki The [https://github.com/unhuman-io/obot OBot] is an open-source robot designed for doing mobile manipulation. {{infobox actuator | name = OBot | voltage = 36V | interface = USB }} 5e2a4717f24711679c0402fd2fa47e818ff8a4f8 0 90 381 2024-04-25T22:20:07Z Ben 2 Created page with "The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. {{infobox actuator | name = SPIN Servo | m..." wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. {{infobox actuator | name = SPIN Servo | manufacturer = | cost = USD 1 | purchase_link = | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = CAN bus | gear_type = }} 9013a7fe37a2c07d94cb3ef217e94dd254cf4ea3 385 381 2024-04-25T22:23:31Z Ben 2 wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. {{infobox actuator | name = SPIN Servo | manufacturer = | cost = USD 30 | purchase_link = | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = CAN bus | gear_type = }} 7681fc84784b2183a0019eab5680467a16ca173a 386 385 2024-04-25T22:39:16Z 157.131.153.166 0 wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque =0.375nm | weight = | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c | gear_type = N/A }} f8ccc0edb7c78ad4dc865f69ea734876d12cc2d0 387 386 2024-04-25T22:40:01Z 157.131.153.166 0 wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque =0.375nm | weight = | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c | gear_type = N/A }} b2fc8d79d30bf4c52c8b49752be9378bfcb73eb2 388 387 2024-04-25T22:43:57Z 157.131.153.166 0 wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque = 0.375nm | weight = 311.6g | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c | gear_type = N/A }} 6f9d1cb38179435da6678448379cb1185fea11b5 390 388 2024-04-25T22:47:23Z Narayanp 12 wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. [[File:Spin.jpg|thumb]] {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque = 0.375nm | weight = 311.6g | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c | gear_type = N/A }} 801ad5ecf42884c4d6aa68565aa766b731c001ce 391 390 2024-04-25T22:59:14Z 24.130.242.94 0 wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. [[File:Spin.jpg|thumb]] {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque = 0.375nm | weight = 311.6g | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c }} 69d4de3f9392354d09974f3447e7b1b76e389ab1 6 91 389 2024-04-25T22:47:16Z Narayanp 12 wikitext text/x-wiki photo of the spin motor and driver board 7e65d2a207ed16ee253b0d74c01af55e054dbe9a 0 16 392 54 2024-04-25T23:14:12Z Ben 2 wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/is_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. 6dd51d68e97e0b5ab26b213c3f5013e13d23639c 0 92 395 2024-04-26T09:28:04Z User2024 6 Created page with "K1 is a humanoid robot from [[Kepler]]. {{infobox robot | name = K1 | organization = [[Kepler]] | height = 178 cm | weight = 85 kg | video_link = https://www.youtube.com/wa..." wikitext text/x-wiki K1 is a humanoid robot from [[Kepler]]. {{infobox robot | name = K1 | organization = [[Kepler]] | height = 178 cm | weight = 85 kg | video_link = https://www.youtube.com/watch?v=A5vshTgDbKE | cost }} [[Category:Robots]] bfc0cbb9186649495e5f82e83ec0db55203ff404 0 93 396 2024-04-26T09:28:29Z User2024 6 Created page with "Kepler is building a humanoid robot called [[K1]]. {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} Ca..." wikitext text/x-wiki Kepler is building a humanoid robot called [[K1]]. {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} [[Category:Companies]] 7827d7a0f8cc8024557ff34624c3bbdd8b007b5d 0 94 397 2024-04-26T09:32:04Z User2024 6 Created page with "Kaleido is a humanoid robot from [[Kawasaki Robotics]]. {{infobox robot | name = Kaleido | organization = [[Kawasaki Robotics]] | height = 179 cm | weight = 83 kg | two_han..." wikitext text/x-wiki Kaleido is a humanoid robot from [[Kawasaki Robotics]]. {{infobox robot | name = Kaleido | organization = [[Kawasaki Robotics]] | height = 179 cm | weight = 83 kg | two_hand_payload = 60 | video_link = https://www.youtube.com/watch?v=_h66xSbIEdU | cost }} [[Category:Robots]] d109d73f886500e4dd0ce2a1b18306c4a00d6085 0 95 398 2024-04-26T09:32:23Z User2024 6 Created page with "Kawasaki Robotics is building a humanoid robot called [[Kaleido]]. {{infobox company | name = Kawasaki Robotics | country = Japan | website_link = https://kawasakirobotics.co..." wikitext text/x-wiki Kawasaki Robotics is building a humanoid robot called [[Kaleido]]. {{infobox company | name = Kawasaki Robotics | country = Japan | website_link = https://kawasakirobotics.com/ | robots = [[Kaleido]] }} [[Category:Companies]] 4c3203aee0a2d49bd2447b02185e2e260905f5ce 400 398 2024-04-26T09:34:19Z User2024 6 wikitext text/x-wiki Kawasaki Robotics is building humanoid robots called [[Kaleido]] and [[Friends]]. {{infobox company | name = Kawasaki Robotics | country = Japan | website_link = https://kawasakirobotics.com/ | robots = [[Kaleido]], [[Friends]] }} [[Category:Companies]] 8b6e10adc633599736ec4b62e4a8a3a788d5dfe7 0 96 399 2024-04-26T09:33:39Z User2024 6 Created page with "Friends is a humanoid robot from [[Kawasaki Robotics]]. {{infobox robot | name = Friends | organization = [[Kawasaki Robotics]] | height = 168 cm | weight = 54 kg | two_han..." wikitext text/x-wiki Friends is a humanoid robot from [[Kawasaki Robotics]]. {{infobox robot | name = Friends | organization = [[Kawasaki Robotics]] | height = 168 cm | weight = 54 kg | two_hand_payload = 10 | video_link = https://www.youtube.com/watch?v=dz4YLbgbVvc | cost }} [[Category:Robots]] 28f51a1d1f16363804e8d537abc3641d19f4af33 0 97 401 2024-04-26T09:37:03Z User2024 6 Created page with "Kangaroo is a humanoid robot from [[PAL Robotics]]. {{infobox robot | name = Kangaroo | organization = [[PAL Robotics]] | height = 145 cm | weight = 40 kg | two_hand_payloa..." wikitext text/x-wiki Kangaroo is a humanoid robot from [[PAL Robotics]]. {{infobox robot | name = Kangaroo | organization = [[PAL Robotics]] | height = 145 cm | weight = 40 kg | two_hand_payload | video_link = https://www.youtube.com/watch?v=TU9q6j8KJGU&t=63s | cost }} [[Category:Robots]] 2a8f82243e0152042b311092c8c18d5bf36fab6b 0 98 402 2024-04-26T09:37:29Z User2024 6 Created page with "PAL Robotics is building a humanoid robot called [[Kangaroo]]. {{infobox company | name = PAL Robotics | country = Spain | website_link = https://pal-robotics.com/ | robots =..." wikitext text/x-wiki PAL Robotics is building a humanoid robot called [[Kangaroo]]. {{infobox company | name = PAL Robotics | country = Spain | website_link = https://pal-robotics.com/ | robots = [[Kangaroo]] }} [[Category:Companies]] a9ffd1d474c012ca8be8a126a748657828473e8c 404 402 2024-04-26T09:40:23Z User2024 6 wikitext text/x-wiki PAL Robotics is building humanoid robots called [[Kangaroo]], [[REEM-C]]. {{infobox company | name = PAL Robotics | country = Spain | website_link = https://pal-robotics.com/ | robots = [[Kangaroo]], [[REEM-C]] }} [[Category:Companies]] f3ca92907f0823f29877e12eb304147fd3c4836c 406 404 2024-04-26T09:41:59Z User2024 6 wikitext text/x-wiki PAL Robotics is building humanoid robots called [[Kangaroo]], [[REEM-C]] and [[TALOS]]. {{infobox company | name = PAL Robotics | country = Spain | website_link = https://pal-robotics.com/ | robots = [[Kangaroo]], [[REEM-C]], [[TALOS]] }} [[Category:Companies]] 25b428f4604c694759e788051599dea01385d853 0 99 403 2024-04-26T09:39:14Z User2024 6 Created page with "REEM-C is a humanoid robot from [[PAL Robotics]]. {{infobox robot | name = REEM-C | organization = [[PAL Robotics]] | height = 165 cm | weight = 80 kg | single_hand_payload..." wikitext text/x-wiki REEM-C is a humanoid robot from [[PAL Robotics]]. {{infobox robot | name = REEM-C | organization = [[PAL Robotics]] | height = 165 cm | weight = 80 kg | single_hand_payload = 1 | two_hand_payload | video_link = https://www.youtube.com/watch?v=lqxTov7isio | cost }} [[Category:Robots]] d931cc751c79a7d21da55638e744ff3106fa4be5 0 100 405 2024-04-26T09:41:33Z User2024 6 Created page with "TALOS is a humanoid robot from [[PAL Robotics]]. {{infobox robot | name = TALOS | organization = [[PAL Robotics]] | height = 175 cm | weight = 95 kg | single_hand_payload =..." wikitext text/x-wiki TALOS is a humanoid robot from [[PAL Robotics]]. {{infobox robot | name = TALOS | organization = [[PAL Robotics]] | height = 175 cm | weight = 95 kg | single_hand_payload = 6 | two_hand_payload | video_link = https://www.youtube.com/watch?v=xUeApfMAKAE | cost }} [[Category:Robots]] 55891c8df21918b1ece73ea95c301a21e695a70e 0 101 407 2024-04-26T09:46:14Z User2024 6 Created page with "Kuavo is a humanoid robot from [[LEJUROBOT ]]. {{infobox robot | name = Kuavo | organization = [[LEJUROBOT]] | height = | weight = 45 kg | single_hand_payload | two_hand_pay..." wikitext text/x-wiki Kuavo is a humanoid robot from [[LEJUROBOT ]]. {{infobox robot | name = Kuavo | organization = [[LEJUROBOT]] | height = | weight = 45 kg | single_hand_payload | two_hand_payload | video_link = https://www.youtube.com/watch?v=Rx1h59y01GY | cost }} [[Category:Robots]] 54fea96758c9519d1c8ddccc0f3decb3d72feff9 0 102 408 2024-04-26T09:46:33Z User2024 6 Created page with "LEJUROBOT is building a humanoid robot called [[Kuavo]]. {{infobox company | name = LEJUROBOT | country = China | website_link = https://www.lejurobot.com/ | robots = Kuav..." wikitext text/x-wiki LEJUROBOT is building a humanoid robot called [[Kuavo]]. {{infobox company | name = LEJUROBOT | country = China | website_link = https://www.lejurobot.com/ | robots = [[Kuavo]] }} [[Category:Companies]] fd19716d4f09956612746a5fb6677549bc08ef95 0 103 409 2024-04-26T09:49:06Z User2024 6 Created page with "MagicBot is a humanoid robot from [[MagicLab, DREAME]]. {{infobox robot | name = MagicBot | organization = [[MagicLab, DREAME]] | height = 178 cm | weight = 56 kg | single_h..." wikitext text/x-wiki MagicBot is a humanoid robot from [[MagicLab, DREAME]]. {{infobox robot | name = MagicBot | organization = [[MagicLab, DREAME]] | height = 178 cm | weight = 56 kg | single_hand_payload | two_hand_payload | video_link = https://www.youtube.com/watch?v=NTPmiDrHv4E&t=2s | cost }} [[Category:Robots]] e881f8e6b334297147f19313b165d13cbe39a98c 0 104 410 2024-04-26T09:49:33Z User2024 6 Created page with "MagicLab, DREAME is building a humanoid robot called [[MagicBot]]. {{infobox company | name = MagicLab, DREAME | country = China | website_link = | robots = [[MagicBot]] }}..." wikitext text/x-wiki MagicLab, DREAME is building a humanoid robot called [[MagicBot]]. {{infobox company | name = MagicLab, DREAME | country = China | website_link = | robots = [[MagicBot]] }} [[Category:Companies]] 44471abf7a6d74abcbd2211e7d3b9ffeb462605d 0 105 411 2024-04-26T09:52:05Z User2024 6 Created page with "MenteeBot (Robot) is a humanoid robot from [[MenteeBot]]. {{infobox robot | name = MenteeBot (Robot) | organization = [[MenteeBot]] | height = 175 cm | weight = 70 kg | sing..." wikitext text/x-wiki MenteeBot (Robot) is a humanoid robot from [[MenteeBot]]. {{infobox robot | name = MenteeBot (Robot) | organization = [[MenteeBot]] | height = 175 cm | weight = 70 kg | single_hand_payload | two_hand_payload = 25 | video_link = https://www.youtube.com/watch?v=zJTf4JhGSsI | cost }} [[Category:Robots]] 703b47fdda3299a6dfb075609150667414b2fb9e 0 106 412 2024-04-26T09:52:18Z User2024 6 Created page with "MenteeBot is building a humanoid robot called [[MenteeBot (Robot)]]. {{infobox company | name = MenteeBot | country = Israel | website_link = https://www.menteebot.com/ | ro..." wikitext text/x-wiki MenteeBot is building a humanoid robot called [[MenteeBot (Robot)]]. {{infobox company | name = MenteeBot | country = Israel | website_link = https://www.menteebot.com/ | robots = [[MenteeBot (Robot)]] }} [[Category:Companies]] d6961a3aab5a6bf59fcdcd5d0ac420d885dfcc71 0 107 413 2024-04-26T09:59:46Z User2024 6 Created page with "Mona is a humanoid robot from [[Kind Humanoid]]. {{infobox robot | name = Mona | organization = [[Kind Humanoid]] | height = | weight = | single_hand_payload | two_hand_pay..." wikitext text/x-wiki Mona is a humanoid robot from [[Kind Humanoid]]. {{infobox robot | name = Mona | organization = [[Kind Humanoid]] | height = | weight = | single_hand_payload | two_hand_payload = | video_link = https://www.youtube.com/watch?v=LsKOpooGK4Q | cost }} [[Category:Robots]] 60d89c2e2198842d1a281125d4b4534881c794a0 0 108 414 2024-04-26T10:00:24Z User2024 6 Created page with "Kind Humanoid is building a humanoid robot called [[Mona]]. {{infobox company | name = Kind Humanoid | country = USA | website_link = https://www.kindhumanoid.com/ | robots =..." wikitext text/x-wiki Kind Humanoid is building a humanoid robot called [[Mona]]. {{infobox company | name = Kind Humanoid | country = USA | website_link = https://www.kindhumanoid.com/ | robots = [[Mona]] }} [[Category:Companies]] 5502863c1226cb565757715388a3c725fb6e2169 0 109 415 2024-04-26T10:02:49Z User2024 6 Created page with "Nadia is a humanoid robot from [[IHMC, Boardwalk Robotics]]. {{infobox robot | name = Nadia | organization = [[IHMC, Boardwalk Robotics]] | height = | weight = | single_han..." wikitext text/x-wiki Nadia is a humanoid robot from [[IHMC, Boardwalk Robotics]]. {{infobox robot | name = Nadia | organization = [[IHMC, Boardwalk Robotics]] | height = | weight = | single_hand_payload | two_hand_payload = | video_link = https://www.youtube.com/watch?v=uTmUfOc7r_s | cost }} [[Category:Robots]] 1a0b3687b268c0ab1789334eac37b6a6736c96a0 0 110 416 2024-04-26T10:02:59Z User2024 6 Created page with "IHMC, Boardwalk Robotics is building a humanoid robot called [[Nadia]]. {{infobox company | name = IHMC, Boardwalk Robotics | country = USA | website_link = https://www.ihmc...." wikitext text/x-wiki IHMC, Boardwalk Robotics is building a humanoid robot called [[Nadia]]. {{infobox company | name = IHMC, Boardwalk Robotics | country = USA | website_link = https://www.ihmc.us/ | robots = [[Nadia]] }} [[Category:Companies]] 507bc2542ec48cd56549666fa8dfd635ff17cfc0 0 111 417 2024-04-26T10:19:25Z User2024 6 Created page with "PX5 is a humanoid robot from [[Xpeng]]. {{infobox robot | name = PX5 | organization = [[Xpeng]] | height = | weight = | single_hand_payload | two_hand_payload = | video_li..." wikitext text/x-wiki PX5 is a humanoid robot from [[Xpeng]]. {{infobox robot | name = PX5 | organization = [[Xpeng]] | height = | weight = | single_hand_payload | two_hand_payload = | video_link = https://www.youtube.com/watch?v=BNSZ8Fwcd20 | cost }} [[Category:Robots]] eb6b4deda29a3f5cdc249dfadfac6f59d9b537ed 0 112 418 2024-04-26T10:19:38Z User2024 6 Created page with "Xpeng is building a humanoid robot called [[PX5]]. {{infobox company | name = Xpeng | country = China | website_link = https://www.xpeng.com/ | robots = [[PX5]] }} Categor..." wikitext text/x-wiki Xpeng is building a humanoid robot called [[PX5]]. {{infobox company | name = Xpeng | country = China | website_link = https://www.xpeng.com/ | robots = [[PX5]] }} [[Category:Companies]] d8b2631a560f611b427ba61fa4f6314ae950e5d3 0 1 419 379 2024-04-26T13:02:59Z 45.27.55.241 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[XPENG]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |} 5b6eb7505c8b3e1d23ff8c471421a0da64d60f2f 0 113 420 2024-04-26T13:07:06Z 45.27.55.241 0 SuperDroid Robots - Building Robots Before It Was Cool wikitext text/x-wiki SuperDroid Robots We’re not your typical robot startup because we’ve been doing this for more than 20 years. Over that time, we’ve built thousands of custom robots, fully assembled robots, and robot kits for nearly every industry and we have clients spanning the globe in over 110 countries. Our mission is to make advanced mobile robots accessible. Our Humanoid product is called [[Rocky]]. {{infobox company | name = SuperDroid Robots | country = USA | website_link = https://superdroidrobots.com/ | robots = [[Rocky]] }} [[Category:Companies]] 0265136212a3eebe0daec003a3f297589628296c 0 114 421 2024-04-26T13:14:41Z 45.27.55.241 0 SuperDroid Robots - Rocky wikitext text/x-wiki Rocky is [[SuperDroid Robots]] most versatile robot platform which can move on any terrain, climb stairs, and step over obstacles. We are planning to open-source the humanoid once it is publicly walking and interacting with physical objects. {{infobox robot | name = Rocky | organization = [[SuperDroid Robots]] | height = 64 in | width= 24 in | length = 12 in | weight = 120 pounds | locomotion control = Deep Reinforcement Learning | software framework = ROS2 | V1 video_link = https://www.youtube.com/watch?v=MvAS4AsMvCI | V2 video_link =https://twitter.com/stevenuecke/status/1707899032973033690 }} [[Category:Robots]] d1a293a5c21242e24ed0b45a5bc81d5a4c5e0add 422 421 2024-04-26T13:15:37Z 45.27.55.241 0 wikitext text/x-wiki Rocky is [[SuperDroid Robots]] most versatile robot platform which can move on any terrain, climb stairs, and step over obstacles. We are planning to open-source the humanoid once it is publicly walking and interacting with physical objects. {{infobox robot | name = Rocky | organization = [[SuperDroid Robots]] | height = 64 in | width= 24 in | length = 12 in | weight = 120 pounds | locomotion control = Deep Reinforcement Learning | software framework = ROS2 | video_link = https://www.youtube.com/watch?v=MvAS4AsMvCI | video_link =https://twitter.com/stevenuecke/status/1707899032973033690 }} [[Category:Robots]] e64c4426305eafc321e9638248fb733fa18edbf6 423 422 2024-04-26T13:20:56Z 45.27.55.241 0 wikitext text/x-wiki Rocky is [[SuperDroid Robots]] most versatile robot platform which can move on any terrain, climb stairs, and step over obstacles. We are planning to open-source the humanoid once it is publicly walking and interacting with physical objects. {{infobox robot | name = Rocky | organization = [[SuperDroid Robots]] | video_link = https://www.youtube.com/watch?v=MvAS4AsMvCI , https://twitter.com/stevenuecke/status/1707899032973033690 | cost = $75,000 | height = 64 in | weight = 120 pounds | speed = | lift_force = 150 lbs | battery_life = 8 hours | battery_capacity = 3,600 Wh | purchase_link = https://www.superdroidrobots.com/humanoid-biped-robot/ | number_made = 1 | dof = 22 | status = Finishing sim-2-real for motors }} [[Category:Robots]] 9601cbc5bbdd23f2eb13887b081a40d1c17bdd7c 424 423 2024-04-26T13:21:39Z 45.27.55.241 0 wikitext text/x-wiki Rocky is [[SuperDroid Robots]] most versatile robot platform which can move on any terrain, climb stairs, and step over obstacles. We are planning to open-source the humanoid once it is publicly walking and interacting with physical objects. {{infobox robot | name = Rocky | organization = [[SuperDroid Robots]] | video_link = V1 https://www.youtube.com/watch?v=MvAS4AsMvCI V2 https://twitter.com/stevenuecke/status/1707899032973033690 | cost = $75,000 | height = 64 in | weight = 120 pounds | speed = | lift_force = 150 lbs | battery_life = 8 hours | battery_capacity = 3,600 Wh | purchase_link = https://www.superdroidrobots.com/humanoid-biped-robot/ | number_made = 1 | dof = 22 | status = Finishing sim-2-real for motors }} [[Category:Robots]] bd104b29463b37fc3e7fde7e2f780ac2da95fce1 14 60 425 300 2024-04-26T18:37:28Z 136.62.52.52 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC, WebRTC, ZMQ. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com * https://github.com/ToruOwO/hato/tree/main?tab=readme-ov-file#collecting-demonstration-data == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 6ee0ac77e73513902bbaa94f5f1cc5743b6c0259 426 425 2024-04-26T18:44:39Z 136.62.52.52 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com * https://github.com/wuphilipp/gello_software == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC, WebRTC, ZMQ. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com * https://github.com/ToruOwO/hato/tree/main?tab=readme-ov-file#collecting-demonstration-data * https://github.com/rail-berkeley/oculus_reader == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] cadfac703cbd83825001628e0799792594708550 427 426 2024-04-26T20:43:30Z 136.62.52.52 0 /* VR Teleop */ wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com * https://github.com/wuphilipp/gello_software == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC, WebRTC, ZMQ. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com * https://github.com/ToruOwO/hato/tree/main?tab=readme-ov-file#collecting-demonstration-data * https://github.com/rail-berkeley/oculus_reader * https://github.com/OpenTeleVision/TeleVision/tree/main == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 79ef74d0e34e2ba9038330d23b8f27b91e43607d 0 115 428 2024-04-27T08:23:36Z User2024 6 Created page with "THEMIS is a humanoid robot from [[Westwood Robotics]]. {{infobox robot | name = THEMIS | organization = [[Westwood Robotics]] | height = 142.2 cm | weight = 39 kg | video_lin..." wikitext text/x-wiki THEMIS is a humanoid robot from [[Westwood Robotics]]. {{infobox robot | name = THEMIS | organization = [[Westwood Robotics]] | height = 142.2 cm | weight = 39 kg | video_link = https://www.youtube.com/watch?v=yt4mHwAl9cc | cost = }} [[Category:Robots]] 6814e6dc170d59ae71cc28f086056ef1a8e19305 0 116 429 2024-04-27T08:23:59Z User2024 6 Created page with "Westwood Robotics is building a humanoid robot called [[THEMIS]]. {{infobox company | name = Westwood Robotics | country = USA | website_link = https://www.westwoodrobotics.i..." wikitext text/x-wiki Westwood Robotics is building a humanoid robot called [[THEMIS]]. {{infobox company | name = Westwood Robotics | country = USA | website_link = https://www.westwoodrobotics.io/ | robots = [[THEMIS]] }} [[Category:Companies]] 39ccc919272594359f47e209bf225c5392b49d6b 0 117 430 2024-04-27T08:26:51Z User2024 6 Created page with "Valkyrie is a humanoid robot from [[NASA]]. {{infobox robot | name = Valkyrie | organization = [[NASA]] | height = 190 cm | weight = 125 kg | video_link = https://www.youtube..." wikitext text/x-wiki Valkyrie is a humanoid robot from [[NASA]]. {{infobox robot | name = Valkyrie | organization = [[NASA]] | height = 190 cm | weight = 125 kg | video_link = https://www.youtube.com/watch?v=LaYlQYHXJio | cost = }} [[Category:Robots]] 7e98af5461e6cd0fbf780e586c061af59c2fc5f6 0 118 431 2024-04-27T08:27:06Z User2024 6 Created page with "NASA is building a humanoid robot called [[Valkyrie]]. {{infobox company | name = NASA | country = USA | website_link = https://www.nasa.gov/ | robots = [[Valkyrie]] }} Ca..." wikitext text/x-wiki NASA is building a humanoid robot called [[Valkyrie]]. {{infobox company | name = NASA | country = USA | website_link = https://www.nasa.gov/ | robots = [[Valkyrie]] }} [[Category:Companies]] d99ef6c93ca6d34f7365fea5ee900f9a2d1fe6ae 0 119 432 2024-04-27T08:30:12Z User2024 6 Created page with "T1 is a humanoid robot from [[FDROBOT]]. {{infobox robot | name = T1 | organization = [[FDROBOT]] | height = 160 cm | weight = 43 kg | single_hand_payload = 145 | video_link..." wikitext text/x-wiki T1 is a humanoid robot from [[FDROBOT]]. {{infobox robot | name = T1 | organization = [[FDROBOT]] | height = 160 cm | weight = 43 kg | single_hand_payload = 145 | video_link = | cost = }} [[Category:Robots]] 20252dac694875d5b758d42d6348970435b549ae 0 120 433 2024-04-27T08:30:22Z User2024 6 Created page with "FDROBOT is building a humanoid robot called [[T1]]. {{infobox company | name = FDROBOT | country = China | website_link = https://twitter.com/FDROBOT192380 | robots = [[T1]]..." wikitext text/x-wiki FDROBOT is building a humanoid robot called [[T1]]. {{infobox company | name = FDROBOT | country = China | website_link = https://twitter.com/FDROBOT192380 | robots = [[T1]] }} [[Category:Companies]] 6250f7f31b0b7d0e1802c739d785a446720155b5 0 121 434 2024-04-27T08:34:33Z User2024 6 Created page with "Figure 01 is a humanoid robot from [[Figure AI]]. {{infobox robot | name = Figure 01 | organization = [[Figure AI]] | height = 167.6 cm | weight = 60 kg | payload = 20 kg | r..." wikitext text/x-wiki Figure 01 is a humanoid robot from [[Figure AI]]. {{infobox robot | name = Figure 01 | organization = [[Figure AI]] | height = 167.6 cm | weight = 60 kg | payload = 20 kg | runtime = 5 Hrs | speed = 1.2 M/s | video_link = https://www.youtube.com/watch?v=48qL8Jt39Vs | cost = }} [[Category:Robots]] ccd62749fab3b126bfe82c3f3a58b315c7b1895b 0 122 435 2024-04-27T08:35:00Z User2024 6 Created page with "Figure AI is building a humanoid robot called [[Figure 01]]. {{infobox company | name = Figure AI | country = USA | website_link = https://www.figure.ai/ | robots = Figure..." wikitext text/x-wiki Figure AI is building a humanoid robot called [[Figure 01]]. {{infobox company | name = Figure AI | country = USA | website_link = https://www.figure.ai/ | robots = [[Figure 01]] }} [[Category:Companies]] 9eb6316809fbb97d1020a3d2532ad1d6ebaebc10 0 123 436 2024-04-27T09:53:57Z User2024 6 Created page with "RAISE-A1 is the First-Generation General Embodied Intelligent Robot by [[AGIBOT]]. {{infobox robot | name = RAISE-A1 | organization = [[AGIBOT]] | height = 175 cm | weight =..." wikitext text/x-wiki RAISE-A1 is the First-Generation General Embodied Intelligent Robot by [[AGIBOT]]. {{infobox robot | name = RAISE-A1 | organization = [[AGIBOT]] | height = 175 cm | weight = 55 kg | single_arm_payload = 5 kg | runtime = 5 Hrs | walk_speed = 7 km/h | video_link = https://www.youtube.com/watch?v=PIYJtZmzs70 | cost = }} [[Category:Robots]] ddef3ec0a8da9773286400faf52429c945f47368 0 124 437 2024-04-27T09:54:20Z User2024 6 Created page with "AGIBOT is building a humanoid robot called [[RAISE-A1]]. {{infobox company | name = AGIBOT | country = China | website_link = https://www.agibot.com/ | robots = [[RAISE-A1]]..." wikitext text/x-wiki AGIBOT is building a humanoid robot called [[RAISE-A1]]. {{infobox company | name = AGIBOT | country = China | website_link = https://www.agibot.com/ | robots = [[RAISE-A1]] }} [[Category:Companies]] ca4ed04cf217b02774b9e97259c37d829d43e63a 0 125 438 2024-04-27T09:58:43Z User2024 6 Created page with "CL-1 is a humanoid robot from [[LimX Dynamics]]. {{infobox robot | name = CL-1 | organization = [[LimX Dynamics]] | height = | weight = | single_arm_payload = | runtime =..." wikitext text/x-wiki CL-1 is a humanoid robot from [[LimX Dynamics]]. {{infobox robot | name = CL-1 | organization = [[LimX Dynamics]] | height = | weight = | single_arm_payload = | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=sihIDeJ4Hmk | cost = }} [[Category:Robots]] 86897dd0dbe52e82677d05c45b7893114a59c8f3 0 126 439 2024-04-27T10:01:10Z User2024 6 Created page with "CyberOne is a humanoid robot from [[Xiaomi]]. {{infobox robot | name = CyberOne | organization = [[Xiaomi]] | height = 177 cm | weight = 52 kg | single_arm_payload = 1.5 | ru..." wikitext text/x-wiki CyberOne is a humanoid robot from [[Xiaomi]]. {{infobox robot | name = CyberOne | organization = [[Xiaomi]] | height = 177 cm | weight = 52 kg | single_arm_payload = 1.5 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=yBmatGQ0giY | cost = }} [[Category:Robots]] 7eb35a02076a9043d82a00ea802e54f0eb1debc0 0 127 440 2024-04-27T10:01:21Z User2024 6 Created page with "Xiaomi is building a humanoid robot called [[CyberOne]]. {{infobox company | name = Xiaomi | country = China | website_link = https://www.mi.com/ | robots = [[CyberOne]] }}..." wikitext text/x-wiki Xiaomi is building a humanoid robot called [[CyberOne]]. {{infobox company | name = Xiaomi | country = China | website_link = https://www.mi.com/ | robots = [[CyberOne]] }} [[Category:Companies]] 3c17c1384231b117f5fb8392eb10caea7e691a90 0 128 441 2024-04-27T10:29:32Z User2024 6 Created page with "Digit is a humanoid robot from [[Agility]]. {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runt..." wikitext text/x-wiki Digit is a humanoid robot from [[Agility]]. {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=NgYo-Wd0E_U | cost = }} [[Category:Robots]] 1b72aeb30ef952d113d6bfcfc087140d5bb5d2d0 450 441 2024-04-27T11:03:28Z User2024 6 wikitext text/x-wiki Digit is a humanoid robot developed by [[Agility]], designed to navigate our world and perform tasks like navigation, obstacle avoidance, and manipulation. It's equipped with a torso full of sensors and a pair of arms, and is considered the most advanced Mobile Manipulation Robot (MMR) on the market, capable of performing repetitive tasks in production environments without requiring significant infrastructure changes. {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=NgYo-Wd0E_U | cost = }} [[Category:Robots]] 8d5847cb50ed4df68f46c5cee9afde1e688e0438 0 129 442 2024-04-27T10:44:46Z User2024 6 Created page with "Draco is a high-performance bipedal platform developed by [[Apptronik]]. It’s their first biped robot, designed with a focus on speed and power. The system has 10 Degrees of..." wikitext text/x-wiki Draco is a high-performance bipedal platform developed by [[Apptronik]]. It’s their first biped robot, designed with a focus on speed and power. The system has 10 Degrees of Freedom (DOFs), allowing for a wide range of movements and tasks. One of the key features of Draco is its liquid cooling system, which helps maintain optimal performance during operation. 9a6591cfae481b160e9b772021a985cd4bb38c1f 443 442 2024-04-27T10:48:34Z User2024 6 wikitext text/x-wiki Draco is a high-performance bipedal platform developed by [[Apptronik]]. It’s their first biped robot, designed with a focus on speed and power. The system has 10 Degrees of Freedom (DOFs), allowing for a wide range of movements and tasks. One of the key features of Draco is its liquid cooling system, which helps maintain optimal performance during operation. Draco is a humanoid robot from [[Apptronik]]. {{infobox robot | name = Draco | organization = [[Apptronik]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=g9xt2zdSOo8 | cost = }} [[Category:Robots]] 48c1093fe79370ca8831605e39e3c4ec293666c7 444 443 2024-04-27T10:48:48Z User2024 6 wikitext text/x-wiki Draco is a high-performance bipedal platform developed by [[Apptronik]]. It’s their first biped robot, designed with a focus on speed and power. The system has 10 Degrees of Freedom (DOFs), allowing for a wide range of movements and tasks. One of the key features of Draco is its liquid cooling system, which helps maintain optimal performance during operation. {{infobox robot | name = Draco | organization = [[Apptronik]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=g9xt2zdSOo8 | cost = }} [[Category:Robots]] dc60b69b0a395093d82a0aef8719264c8f89fcbb 0 130 445 2024-04-27T10:51:16Z User2024 6 Created page with "Apptronik is building humanoid robot called [[Draco]]. {{infobox company | name = Apptronik | country = USA | website_link = https://apptronik.com/ | robots = [[Draco]] }} [..." wikitext text/x-wiki Apptronik is building humanoid robot called [[Draco]]. {{infobox company | name = Apptronik | country = USA | website_link = https://apptronik.com/ | robots = [[Draco]] }} [[Category:Companies]] 26e0d333e5bcd6da2ff79675cf94db76bf96b974 446 445 2024-04-27T10:52:03Z User2024 6 wikitext text/x-wiki Apptronik is building humanoid robots called [[Draco]], [[Valkyrie]]. {{infobox company | name = Apptronik | country = USA | website_link = https://apptronik.com/ | robots = [[Draco]], [[Valkyrie]] }} [[Category:Companies]] 74a5a9dd9d000a186c31e2f483de0f3f4ddc7e21 449 446 2024-04-27T11:00:39Z User2024 6 wikitext text/x-wiki Apptronik is building humanoid robots called [[Draco]], [[Valkyrie]], [[Apollo]]. {{infobox company | name = Apptronik | country = USA | website_link = https://apptronik.com/ | robots = [[Draco]], [[Valkyrie]], [[Apollo]] }} [[Category:Companies]] cd53fe3ca7e563003917e2091ddd8e552708b402 0 131 447 2024-04-27T10:59:44Z User2024 6 Created page with "The humanoid robot Apollo is a creation of [[Apptronik]], a company known for its advanced humanoid robots. Apollo is a practical bipedal platform that’s designed to perform..." wikitext text/x-wiki The humanoid robot Apollo is a creation of [[Apptronik]], a company known for its advanced humanoid robots. Apollo is a practical bipedal platform that’s designed to perform useful tasks. It’s equipped with two NVIDIA Jetson units and has been trained in the Isaac platform’s simulation environment. {{infobox robot | name = Draco | organization = [[Apptronik]] | height = 172.7 cm | weight = 73 kg | two_hand_payload = 25 | video_link = https://www.youtube.com/watch?v=3CdwPGC9nyk&t=6s | cost = }} [[Category:Robots]] e68696fb8a554981a2f7d15c305c458f82628815 448 447 2024-04-27T11:00:14Z User2024 6 wikitext text/x-wiki The humanoid robot Apollo is a creation of [[Apptronik]], a company known for its advanced humanoid robots. Apollo is a practical bipedal platform that’s designed to perform useful tasks. It’s equipped with two NVIDIA Jetson units and has been trained in the Isaac platform’s simulation environment. {{infobox robot | name = Apollo | organization = [[Apptronik]] | height = 172.7 cm | weight = 73 kg | two_hand_payload = 25 | video_link = https://www.youtube.com/watch?v=3CdwPGC9nyk&t=6s | cost = }} [[Category:Robots]] 9fbb1c97e4b4ca35032532aad06f947dfc10f962 0 54 451 245 2024-04-27T11:10:49Z User2024 6 wikitext text/x-wiki EVE is a versatile and agile humanoid robot developed by [[1X]]. It is equipped with cameras and sensors to perceive and interact with its surroundings. EVE’s mobility, dexterity, and balance allow it to navigate complex environments and manipulate objects effective. {{infobox robot | name = EVE | organization = [[1X]] | height = 186 cm | weight = 86 kg | speed = 14.4 km/hr | carry_capacity = 15 kg | runtime = 6 hrs | video_link = https://www.youtube.com/watch?v=20GHG-R9eFI }} [[Category:Robots]] ac4895a1e3c3b9edc77b378bc670bc67c22666ae 0 55 452 246 2024-04-27T11:14:50Z User2024 6 wikitext text/x-wiki NEO is a bipedal humanoid robot developed by [[1X]]. It is designed to look and move like a human, featuring a head, torso, arms, and legs. NEO can perform a wide range of tasks, excelling in industrial sectors like security, logistics, manufacturing, operating machinery, and handling complex tasks. It is also envisioned to provide valuable home assistance and perform chores like cleaning or organizing. {{infobox robot | name = NEO | organization = [[1X]] | height = 165 cm | weight = 30 kg | video_link = https://www.youtube.com/watch?v=ikg7xGxvFTs | speed = 4 km/hr | carry_capacity = 20 kg | runtime = 2-4 hrs }} [[Category:Robots]] ef6ec11b98a7c785b7c3d3b46ca01e02a6170672 0 57 453 250 2024-04-27T11:17:38Z User2024 6 wikitext text/x-wiki GR-1 is a self-developed and mass-produced humanoid robot by [[Fourier Intelligence]]. It has a highly bionic torso and human-like motion control capabilities, with up to 54 Degrees of Freedom (DoFs) across its form. GR-1 can walk briskly, adroitly avoid obstacles, stably descend a slope, and withstand disruptions, making it an ideal physical agent of artificial general intelligence (AGI). {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | height = 165 cm | weight = 55 kg | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 | cost = USD 149,999 }} [[Category:Robots]] 46362f2ec92e2bfe21cb5db031780b2567c58cb8 0 71 454 313 2024-04-27T12:09:15Z User2024 6 wikitext text/x-wiki Walker X is a highly advanced AI humanoid robot developed by [[UBTECH]]. It incorporates six cutting-edge AI technologies, including upgraded vision-based navigation and hand-eye coordination, enabling it to move smoothly and quickly, and engage in precise and safe interactions. it is equipped with 41 high-performance servo joints, a 160° face surrounding 4.6K HD dual flexible curved screen, and a 4-dimensional light language system. {{infobox robot | name = Walker X | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1.5 | two_hand_payload = 3 | cost = USD 960,000 | video_link = https://www.youtube.com/watch?v=4ZL3LgdKNbw }} [[Category:Robots]] 7b482c461d06f80601a6dddcdbda0adba80483f3 0 73 455 314 2024-04-27T12:12:57Z User2024 6 wikitext text/x-wiki Panda Robot by [[UBTECH]] is a humanoid robot that was created using the iconic panda image and includes original cutting-edge technologies based on the humanoid service robot, Walker1. This robot has an expressive and life-like expression along with a multi-modal design. {{infobox robot | name = Panda Robot | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1.5 | two_hand_payload = 3 | cost = USD 960,000 }} [[Category:Robots]] 850368072fbb1748ed46620f5e9dee7be12e96ef 0 74 456 317 2024-04-27T12:17:20Z User2024 6 wikitext text/x-wiki Walker S by [[UBTech]] is a highly advanced humanoid robot designed to serve in household and office scenarios. It is equipped with 36 high-performance servo joints and a full range of sensory systems including force, vision, hearing, and spatial awareness, enabling smooth and fast walking and flexible, precise handling. {{infobox robot | name = Walker S | organization = [[UBTech]] | height = | weight = | video_link = https://www.youtube.com/watch?v=UCt7qPpTt-g | single_hand_payload = | two_hand_payload = | cost = }} [[Category:Robots]] 29aa768cbcbeb88a8767dd7274790b1e352120c6 0 75 457 319 2024-04-27T12:20:08Z User2024 6 wikitext text/x-wiki Wukong-IV is an adult-size humanoid robot designed and built by the research team at [[Deep Robotics]]. It is actuated by electric motor joints. The robot has 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm. {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_payload | two_hand_payload | cost = | video_link = https://www.youtube.com/watch?v=fbk4fYc6U14 }} [[Category:Robots]] 44f00af73326d5a6a391a91e6668a939c10a41bc 0 132 458 2024-04-27T12:26:17Z User2024 6 Created page with "XBot is a humanoid robot developed by [[Robot Era]], a startup incubated by Tsinghua University. The company has open-sourced its reinforcement learning framework, Humanoid-Gy..." wikitext text/x-wiki XBot is a humanoid robot developed by [[Robot Era]], a startup incubated by Tsinghua University. The company has open-sourced its reinforcement learning framework, Humanoid-Gym, which was used to train the XBot and has proven successful in sim-to-real policy transfer. {{infobox robot | name = Xbot | organization = [[Robot Era]] | height = 122 cm | weight = 38 kg | two_hand_payload = 25 | video_link = https://www.youtube.com/watch?v=4tiVkZBw188 | cost = }} [[Category:Robots]] ff6bbf459ccf205b3a714089b5444310c1d21283 0 133 459 2024-04-27T12:26:48Z User2024 6 Created page with "Robot Era is building humanoid robot called [[Xbot]]. {{infobox company | name = Robot Era | country = China | website_link = https://www.robotera.com/ | robots = [[Xbot]] }}..." wikitext text/x-wiki Robot Era is building humanoid robot called [[Xbot]]. {{infobox company | name = Robot Era | country = China | website_link = https://www.robotera.com/ | robots = [[Xbot]] }} [[Category:Companies]] 3fc5c050e30e391b5821a1fa6420df9ad3df923e 0 77 460 321 2024-04-27T12:35:56Z User2024 6 wikitext text/x-wiki The XR4, also known as one of the Seven Fairies named Xiaozi, is a humanoid bipedal robot developed by [[Dataa Robotics]]. XR4 is made from lightweight high-strength carbon fiber composite material with over 60 intelligent flexible joints. {{infobox robot | name = XR4 | organization = [[DATAA Robotics]] | height = 168 cm | weight = 65 kg | video_link = https://www.youtube.com/watch?v=DUyUZcH5uUU | single_hand_payload | two_hand_payload | cost = }} [[Category:Robots]] 18a87745e6f5d01e9ce4b87b63fb54b84b3f25d8 0 79 461 323 2024-04-27T14:44:02Z User2024 6 wikitext text/x-wiki ZEUS2Q is a humanoid robot developed by [[System Technology Works]]. It’s a stand-alone system that harnesses the power of Edge AI computing, enabling it to perform localized AI tasks like communication, facial, and object recognition on the edge. {{infobox robot | name = ZEUS2Q | organization = [[System Technology Works]] | height = 127 cm | weight = 13.61 kg | video_link = https://www.youtube.com/watch?v=eR2HMykMITY }} [[Category:Robots]] e43d879da7b8f0f92523c99ae53862afb9b3508a 0 134 462 2024-04-27T14:48:04Z User2024 6 Created page with "HD Atlas is a highly dynamic humanoid robot developed by [[Boston Dynamics]]. It’s designed for real-world applications and is capable of demonstrating advanced athletics an..." wikitext text/x-wiki HD Atlas is a highly dynamic humanoid robot developed by [[Boston Dynamics]]. It’s designed for real-world applications and is capable of demonstrating advanced athletics and agility. {{infobox robot | name = HD Atlas | organization = [[Boston Dynamics]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=-9EM5_VFlt8 | cost = }} [[Category:Robots]] 639b2a2c72fa931b7b173eb40bb2b73f3e1b6218 0 82 463 329 2024-04-27T14:49:05Z User2024 6 wikitext text/x-wiki Boston Dynamics is building humanoid robots called [[Atlas]] and [[HD Atlas]]. {{infobox company | name = Boston Dynamics | country = USA | website_link = https://bostondynamics.com/ | robots = [[Atlas]], [[HD Atlas]] }} [[Category:Companies]] f585f5da86cc0e5a8bffe4ba99e3742a9b8e58cd 0 117 464 430 2024-04-27T14:51:10Z User2024 6 wikitext text/x-wiki [[NASA]]’s Valkyrie, also known as R5, is a robust, rugged, and entirely electric humanoid robot. It was designed and built by the Johnson Space Center (JSC) Engineering Directorate to compete in the 2013 DARPA Robotics Challenge (DRC) Trials. {{infobox robot | name = Valkyrie | organization = [[NASA]] | height = 190 cm | weight = 125 kg | video_link = https://www.youtube.com/watch?v=LaYlQYHXJio | cost = }} [[Category:Robots]] d563055d5de103ac4c32f7700cc28c14d142db7a 465 464 2024-04-27T14:54:59Z User2024 6 Blanked the page wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 466 465 2024-04-27T14:55:31Z User2024 6 wikitext text/x-wiki NASA’s Valkyrie, also known as R5, is a robust, rugged, and entirely electric humanoid robot. It was designed and built by the Johnson Space Center (JSC) Engineering Directorate to compete in the 2013 DARPA Robotics Challenge (DRC) Trials. {{infobox robot | name = Valkyrie | organization = [[NASA]] | height = 190 cm | weight = 125 kg | video_link = https://www.youtube.com/watch?v=LaYlQYHXJio | cost = }} [[Category:Robots]] fa7979a3fc124ab7a4b2a4b28b2a912b541e946b 0 123 467 436 2024-04-27T15:01:58Z User2024 6 wikitext text/x-wiki RAISE-A1 is the first-generation general embodied intelligent robot developed by [[AGIBOT]]. The robot showcases industry-leading capabilities in bipedal walking intelligence and human-machine interaction, and is designed for use in various fields such as flexible manufacturing, interactive services, education and research, specialized avatars, warehousing logistics, and robotic household assistants. {{infobox robot | name = RAISE-A1 | organization = [[AGIBOT]] | height = 175 cm | weight = 55 kg | single_arm_payload = 5 kg | runtime = 5 Hrs | walk_speed = 7 km/h | video_link = https://www.youtube.com/watch?v=PIYJtZmzs70 | cost = }} [[Category:Robots]] 1a3f05eb51515f3a8bb23b9084f998168504b394 0 125 468 438 2024-04-27T15:04:02Z User2024 6 wikitext text/x-wiki The CL-1 is a humanoid robot developed by [[LimX Dynamics]]. It’s one of the few robots globally that can dynamically climb stairs based on real-time terrain perception. This advanced capability is attributed to [[LimX Dynamics]]’ motion control and AI algorithms, as well as its proprietary high-performing actuators and hardware system. {{infobox robot | name = CL-1 | organization = [[LimX Dynamics]] | height = | weight = | single_arm_payload = | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=sihIDeJ4Hmk | cost = }} [[Category:Robots]] 215e104b574859dfa8dda7f1db0b17121e13c723 0 135 469 2024-04-27T15:09:15Z User2024 6 Created page with "The T-HR3 is the third generation humanoid robot unveiled by [[Toyota]]’s Partner Robot Division. It’s designed to explore new technologies for safely managing physical in..." wikitext text/x-wiki The T-HR3 is the third generation humanoid robot unveiled by [[Toyota]]’s Partner Robot Division. It’s designed to explore new technologies for safely managing physical interactions between robots and their surroundings, and it features a remote maneuvering system that mirrors user movements to the robot. The T-HR3 can assist humans in various settings such as homes, medical facilities, construction sites, disaster-stricken areas, and even outer space. {{infobox robot | name = T-HR3 | organization = [[Toyota]] | height = 150 cm | weight = 75 kg | two_hand_payload = | video_link = https://www.youtube.com/watch?v=5dPY7l7u_z0 | cost = }} [[Category:Robots]] 692e0548015f7b26cb33dad8f97753d73bd37c41 472 469 2024-04-27T15:15:11Z User2024 6 wikitext text/x-wiki The T-HR3 is the third generation humanoid robot unveiled by [[Toyota Research Institute]]. It’s designed to explore new technologies for safely managing physical interactions between robots and their surroundings, and it features a remote maneuvering system that mirrors user movements to the robot. The T-HR3 can assist humans in various settings such as homes, medical facilities, construction sites, disaster-stricken areas, and even outer space. {{infobox robot | name = T-HR3 | organization = [[Toyota Research Institute]] | height = 150 cm | weight = 75 kg | two_hand_payload = | video_link = https://www.youtube.com/watch?v=5dPY7l7u_z0 | cost = }} [[Category:Robots]] 90922e534c0b2d82ee92eba63498f0bacef89a19 473 472 2024-04-27T15:15:50Z User2024 6 wikitext text/x-wiki The T-HR3 is the third generation humanoid robot unveiled by [[Toyota]]. It’s designed to explore new technologies for safely managing physical interactions between robots and their surroundings, and it features a remote maneuvering system that mirrors user movements to the robot. The T-HR3 can assist humans in various settings such as homes, medical facilities, construction sites, disaster-stricken areas, and even outer space. {{infobox robot | name = T-HR3 | organization = [[Toyota Research Institute]] | height = 150 cm | weight = 75 kg | two_hand_payload = | video_link = https://www.youtube.com/watch?v=5dPY7l7u_z0 | cost = }} [[Category:Robots]] ce981df64a53c077dd5d039e9722b58d0b51eb56 475 473 2024-04-27T15:16:48Z User2024 6 wikitext text/x-wiki The T-HR3 is the third generation humanoid robot unveiled by [[Toyota Research Institute]]. It’s designed to explore new technologies for safely managing physical interactions between robots and their surroundings, and it features a remote maneuvering system that mirrors user movements to the robot. The T-HR3 can assist humans in various settings such as homes, medical facilities, construction sites, disaster-stricken areas, and even outer space. {{infobox robot | name = T-HR3 | organization = [[Toyota Research Institute]] | height = 150 cm | weight = 75 kg | two_hand_payload = | video_link = https://www.youtube.com/watch?v=5dPY7l7u_z0 | cost = }} [[Category:Robots]] 90922e534c0b2d82ee92eba63498f0bacef89a19 0 136 470 2024-04-27T15:10:26Z User2024 6 Created page with "Toyota is building humanoid robot called [[T-HR3]]. {{infobox company | name = Toyota | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} Cat..." wikitext text/x-wiki Toyota is building humanoid robot called [[T-HR3]]. {{infobox company | name = Toyota | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} [[Category:Companies]] 318959cdb49e082c001c3a40301f8ce6533a52a7 474 470 2024-04-27T15:16:27Z User2024 6 wikitext text/x-wiki Toyota is building humanoid robot called [[T-HR3]]. {{infobox company | name = Toyota Research Institute | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} [[Category:Companies]] 88e5773390dc3ff8b7ff1d3b7a7e0628aa22a2c6 0 137 471 2024-04-27T15:14:11Z User2024 6 Created page with "Punyo is a soft robot developed by the [[Toyota Research Institute]] (TRI) to revolutionize whole-body manipulation research. Unlike traditional robots that primarily use hand..." wikitext text/x-wiki Punyo is a soft robot developed by the [[Toyota Research Institute]] (TRI) to revolutionize whole-body manipulation research. Unlike traditional robots that primarily use hands for manipulation, Punyo employs its arms and chest. The robot is designed to help with everyday tasks, such as lifting heavy objects or closing a drawer. {{infobox robot | name = Punyo | organization = [[Toyota Research Institute]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=FY-MD4gteeE | cost = }} [[Category:Robots]] fdf2d1df61931bb28cf62592bacae471e9973899 0 138 476 2024-04-27T15:17:17Z User2024 6 Created page with "Toyota Research Institute is building humanoid robot called [[T-HR3]]. {{infobox company | name = Toyota Research Institute | country = Japan | website_link = https://global...." wikitext text/x-wiki Toyota Research Institute is building humanoid robot called [[T-HR3]]. {{infobox company | name = Toyota Research Institute | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} [[Category:Companies]] ff87bdcc0c895e7d79b56332d5b000a4a01b3596 0 139 477 2024-04-27T20:28:02Z Ben 2 Created page with "Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Pre-Internship Checklist === - Create a Wiki account - Add yourself to K-Scale E..." wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Pre-Internship Checklist === - Create a Wiki account - Add yourself to [[K-Scale Employees]] e38ee821ad3944b8781f71b270c88ac1aaeda29c 478 477 2024-04-27T20:28:09Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Pre-Internship Checklist === * Create a Wiki account * Add yourself to [[K-Scale Employees]] bbbddfd920caf01d4ffaccf2eee452879439b31b 482 478 2024-04-27T20:32:29Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Pre-Internship Checklist === * Create a wiki account and mark yourself as an employee (you can use [[User:Ben]] as a template). You'll use your account as the main way to keep track of what you've done over the course of the internship. * Contribute an article about something you find interesting. See the [[Contributing]] guide. 4944d70be5b5a3f2fb3f2e126bc379cd6f0bcb67 0 140 479 2024-04-27T20:28:53Z Ben 2 Created page with "This is an unofficial list of all current and past employees. === Current Employees === * [[User:Ben]]" wikitext text/x-wiki This is an unofficial list of all current and past employees. === Current Employees === * [[User:Ben]] 3e0461d5fa7adca5f18bc570900fe4beebd2d85c 2 141 480 2024-04-27T20:29:24Z Ben 2 Created page with "[https://ben.bolte.cc/ Ben] is the founder and CEO of [[K-Scale Labs]]. [[Category: K-Scale Employees]]" wikitext text/x-wiki [https://ben.bolte.cc/ Ben] is the founder and CEO of [[K-Scale Labs]]. [[Category: K-Scale Employees]] cac9f0ee540eea47256394c9cfe52a5cc8fbe253 485 480 2024-04-27T20:43:28Z Ben 2 wikitext text/x-wiki [https://ben.bolte.cc/ Ben] is the founder and CEO of [[K-Scale Labs]]. {{infobox person | name = Ben Bolte | organization = [[K-Scale Labs]] | title = CEO | website_link = https://ben.bolte.cc/ }} [[Category: K-Scale Employees]] a5db4124c07ac2c71491cd255b7cc7e15d338a7d 488 485 2024-04-27T20:47:22Z Ben 2 wikitext text/x-wiki [[File:Ben.jpg|right|200px|thumb]] [https://ben.bolte.cc/ Ben] is the founder and CEO of [[K-Scale Labs]]. {{infobox person | name = Ben Bolte | organization = [[K-Scale Labs]] | title = CEO | website_link = https://ben.bolte.cc/ }} [[Category: K-Scale Employees]] f971d4004fa84319343a24548245ebcf11a88d16 14 142 481 2024-04-27T20:29:58Z Ben 2 Created page with "This category is for users who self-identify as employees at [[K-Scale Labs]]." wikitext text/x-wiki This category is for users who self-identify as employees at [[K-Scale Labs]]. 41ce1a8090e42fc0c7dfe8ce65a65dc2154608ed 10 143 483 2024-04-27T20:42:00Z Ben 2 Created page with "{{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Website | value2 = {{#if: {{{website_link|}}} | [{{{website_link}}} Website] }} }}" wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Website | value2 = {{#if: {{{website_link|}}} | [{{{website_link}}} Website] }} }} f2aa4d7f7697d8ccdf721ffa685ab27118297fbc 484 483 2024-04-27T20:43:24Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Title | value3 = {{{title|}}} | key4 = Website | value4 = {{#if: {{{website_link|}}} | [{{{website_link}}} Website] }} }} 101094484a12137b202efeac2de5a495b37e942f 6 144 486 2024-04-27T20:44:44Z ThanatosRobo 13 wikitext text/x-wiki Concept Render of Neo 65a6e2d609f46bdcaa86e6bc123df569d52693cb 6 145 487 2024-04-27T20:47:12Z Ben 2 wikitext text/x-wiki Ben and his wife next to a rocket d3be019551b44d5fdf287bb2cbd3867095f29680 0 16 489 392 2024-04-27T20:50:31Z Kewang 11 /* Onboarding */ wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. d3af9079c5c5eb43a61f0d20967a74be25564fce 490 489 2024-04-27T20:54:32Z Kewang 11 /* Onboarding */ wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at ~/.ssh/config for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. 368dc849f1da29eae210a5e0656b854e3a7937e8 491 490 2024-04-27T20:56:24Z Kewang 11 /* Onboarding */ wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config<code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. 555e0b7bdd88366179ef8067fd0d4f96ddfb7a36 492 491 2024-04-27T20:56:45Z Kewang 11 /* Onboarding */ wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. f560b17e0ba5c9f13f967e5e6d9c90d779be7572 500 492 2024-04-27T21:10:08Z Kewang 11 /* Onboarding */ wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> You need to restart ssh to get it working. After setting this up, you can use the command <code>ssh cluster</code> to directly connect. <br> You can also access via VScode. Tutorial of using ssh in VScode is [https://code.visualstudio.com/docs/remote/ssh-tutorial Here].<br> Please inform us if you have any issues. === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. cf5db2b85f16a040c7758240f7f3e973f4a9c0fd 501 500 2024-04-27T21:11:26Z Stompy 14 wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> You need to restart <code>ssh</code> to get it working. After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. 9dc6c4c41600329bc73427dfe7b564dc4fbacb72 502 501 2024-04-27T21:11:43Z Stompy 14 wikitext text/x-wiki The K-Scale Labs cluster is a shared cluster for robotics research. This page contains notes on how to access the cluster. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. 6047589d3e88af7467e2203e2abd7c212967845f 0 55 493 452 2024-04-27T21:02:34Z ThanatosRobo 13 wikitext text/x-wiki [[File:1x Neo.jpg|thumb]] NEO is a bipedal humanoid robot developed by [[1X]]. It is designed to look and move like a human, featuring a head, torso, arms, and legs. NEO can perform a wide range of tasks, excelling in industrial sectors like security, logistics, manufacturing, operating machinery, and handling complex tasks. It is also envisioned to provide valuable home assistance and perform chores like cleaning or organizing. NEO's soft, tendon based design is meant to have very low inertia, intended to work in close proximity to humans. It will weigh 30 kilograms, with a 20 kilogram carrying capacity. 1X hopes for Neo to be "an all-purpose android assistant to your daily life." {{infobox robot | name = NEO | organization = [[1X]] | height = 165 cm | weight = 30 kg | video_link = https://www.youtube.com/watch?v=ikg7xGxvFTs | speed = 4 km/hr | carry_capacity = 20 kg | runtime = 2-4 hrs }} [[Category:Robots]] == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. 8493ae5291a723660229ccc6cda4a5cdcb49dfdf 494 493 2024-04-27T21:03:06Z ThanatosRobo 13 wikitext text/x-wiki [[File:1x Neo.jpg|thumb]] {{infobox robot | name = NEO | organization = [[1X]] | height = 165 cm | weight = 30 kg | video_link = https://www.youtube.com/watch?v=ikg7xGxvFTs | speed = 4 km/hr | carry_capacity = 20 kg | runtime = 2-4 hrs }} NEO is a bipedal humanoid robot developed by [[1X]]. It is designed to look and move like a human, featuring a head, torso, arms, and legs. NEO can perform a wide range of tasks, excelling in industrial sectors like security, logistics, manufacturing, operating machinery, and handling complex tasks. It is also envisioned to provide valuable home assistance and perform chores like cleaning or organizing. NEO's soft, tendon based design is meant to have very low inertia, intended to work in close proximity to humans. It will weigh 30 kilograms, with a 20 kilogram carrying capacity. 1X hopes for Neo to be "an all-purpose android assistant to your daily life." [[Category:Robots]] == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. 74952b022ae0142020baeb20bcce032db991611f 6 146 495 2024-04-27T21:06:59Z ThanatosRobo 13 wikitext text/x-wiki A profile view of Digit, from Agility Robotics. 73258e44e0abebd8e828ecf575ab6ad127a898b3 498 495 2024-04-27T21:08:28Z ThanatosRobo 13 wikitext text/x-wiki A frontal view of Digit, from Agility Robotics. 01a5eafca0e50652afc74e15364669155a8409a1 0 7 496 171 2024-04-27T21:07:33Z Ben 2 wikitext text/x-wiki Tesla is building a humanoid robot called [[Optimus]]. {{infobox company | name = Tesla | country = United States | website_link = https://www.tesla.com/ | robots = [[Optimus]] }} [[Category:Companies]] [[Category: Stompy, Expand!]] 4fcc635104d2164cc72670274ac6f743dcae6159 504 496 2024-04-27T22:23:22Z Stompy 14 Bot expanded article wikitext text/x-wiki Tesla, an innovative company known for its electric vehicles and sustainable energy solutions, has embarked on the development of a humanoid robot named Optimus. Optimus is envisioned to be a groundbreaking humanoid robot designed by Tesla to perform a wide range of tasks autonomously. Tesla's foray into humanoid robotics with Optimus showcases the company's commitment to advancing technology beyond the realm of electric vehicles. This project signifies Tesla's exploration of robotics and artificial intelligence beyond its traditional automotive focus. By venturing into humanoid robot development, Tesla aims to revolutionize industries and everyday tasks with cutting-edge automation. For more information about Tesla and its humanoid robot Optimus, interested individuals can visit Tesla's official website. {{infobox company | name = Tesla | country = United States | website_link = https://www.tesla.com/ | robots = [[Optimus]] }} [[Category:Companies]] [[Category: Robotics]] c7e117907014fcd2b2d2a76f3d6b4b41d0ae9cf6 505 504 2024-04-27T22:23:44Z Stompy 14 wikitext text/x-wiki Tesla, an innovative company known for its electric vehicles and sustainable energy solutions, has embarked on the development of a humanoid robot named Optimus. Optimus is envisioned to be a groundbreaking humanoid robot designed by Tesla to perform a wide range of tasks autonomously. Tesla's foray into humanoid robotics with Optimus showcases the company's commitment to advancing technology beyond the realm of electric vehicles. This project signifies Tesla's exploration of robotics and artificial intelligence beyond its traditional automotive focus. By venturing into humanoid robot development, Tesla aims to revolutionize industries and everyday tasks with cutting-edge automation. For more information about Tesla and its humanoid robot Optimus, interested individuals can visit Tesla's official website. {{infobox company | name = Tesla | country = United States | website_link = https://www.tesla.com/ | robots = [[Optimus]] }} [[Category: Companies]] [[Category: Robotics]] 53781cc2b9abffb320b2a5b310b01e5a469bff98 0 128 497 450 2024-04-27T21:07:34Z ThanatosRobo 13 wikitext text/x-wiki [[File:Agility Robotics Digit.jpg|thumb]] {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=NgYo-Wd0E_U | cost = }} Digit is a humanoid robot developed by [[Agility]], designed to navigate our world and perform tasks like navigation, obstacle avoidance, and manipulation. It's equipped with a torso full of sensors and a pair of arms, and is considered the most advanced Mobile Manipulation Robot (MMR) on the market, capable of performing repetitive tasks in production environments without requiring significant infrastructure changes. Digit is notably designed to be bipedal, but not necessarily humanoid/anthropomorphic, with ostrich-like reverse jointed legs. This is a side effect of Agility's design goals, to maximize efficiency and robustness of legged locomotion. [[Category:Robots]] == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. d5d6ac097677e4ef87cd3528e04571054b0beaa1 508 497 2024-04-27T22:25:18Z Stompy 14 wikitext text/x-wiki [[File:Agility Robotics Digit.jpg|thumb]] {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=NgYo-Wd0E_U | cost = }} Digit is a humanoid robot developed by [[Agility]], designed to navigate our world and perform tasks like navigation, obstacle avoidance, and manipulation. It's equipped with a torso full of sensors and a pair of arms, and is considered the most advanced Mobile Manipulation Robot (MMR) on the market, capable of performing repetitive tasks in production environments without requiring significant infrastructure changes. Digit is notably designed to be bipedal, but not necessarily humanoid/anthropomorphic, with ostrich-like reverse jointed legs. This is a side effect of Agility's design goals, to maximize efficiency and robustness of legged locomotion. == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. [[Category:Robots]] [[Category:Stompy, Expand!]] 656205978bebdca6f36a77bbe24a0e1cb8b6721e 512 508 2024-04-27T22:26:03Z Stompy 14 Bot expanded article wikitext text/x-wiki {{Infobox robot | name = Digit | organization = [[Agility Robotics]] | height = 175.3 cm | weight = 65 kg | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} Digit is a bipedal humanoid robot created by Agility Robotics. It stands at 175.3 cm tall and weighs 65 kg. Digit is designed to operate in real-world environments, undertaking tasks such as navigation, obstacle avoidance, and manipulation. The robot is equipped with a variety of sensors along with a pair of arms. It is considered one of the most advanced Mobile Manipulation Robots (MMRs) available, capable of executing repetitive tasks in production settings with minimal need for alterations to existing infrastructure. Unlike traditional humanoid robots, Digit features ostrich-like reverse-jointed legs, reflecting Agility Robotics' focus on optimizing efficiency and robustness in legged locomotion. == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work." ''IEEE Spectrum''. [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Agility Robotics]] 71800de3d0d22a4476f84e5e72ad20b71a82d369 2 147 499 2024-04-27T21:09:35Z Stompy 14 Created page with "I am a bot created by [[K-Scale Labs]]." wikitext text/x-wiki I am a bot created by [[K-Scale Labs]]. 8aff304e49714aef1eb07ac277110d3e055432c6 14 148 503 2024-04-27T21:13:17Z Stompy 14 Created page with "Adding a page to this category will tell Stompy to expand on the information in the page." wikitext text/x-wiki Adding a page to this category will tell Stompy to expand on the information in the page. 2d47af696ce972fa5deeb524d6dc6c85d54e4374 0 53 506 337 2024-04-27T22:24:57Z Stompy 14 wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] [[Category:Robots]] [[Category:Stompy, Expand!]] 189bacf739840735b0f35352a86fdcb767a7e98a 0 9 507 215 2024-04-27T22:25:05Z Stompy 14 wikitext text/x-wiki Sanctuary AI is a humanoid robot company. Their robot is called [[Phoenix]]. {{infobox company | name = Sanctuary | country = United States | website_link = https://sanctuary.ai/ | robots = [[Phoenix]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 3863045c10f1142e80d0fc3fa0f991a98a24308b 0 8 509 192 2024-04-27T22:25:35Z Stompy 14 wikitext text/x-wiki Agility has built several robots. Their humanoid robot is called [[Digit]]. {{infobox company | name = Agility | country = United States | website_link = https://agilityrobotics.com/ | robots = [[Cassie]], [[Digit]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 7dc90d05a7ef4e4d507732655b38676ae20c0474 511 509 2024-04-27T22:25:59Z Stompy 14 Bot expanded article wikitext text/x-wiki Agility is an American robotics company known for designing and manufacturing advanced robots. Among their creations, Agility has developed the humanoid robots named Cassie and Digit. Digit, one of Agility's notable robots, is specifically designed as a humanoid robot. It is known for its advanced bipedal locomotion capabilities and dexterous manipulation abilities. Agility's continued innovation in the field of robotics, particularly in the development of humanoid robots like Digit, highlights the company's commitment to pushing the boundaries of robotic technology. For more information about Agility and their robots, you can visit their official website. {{infobox company | name = Agility | country = United States | website_link = https://agilityrobotics.com/ | robots = [[Cassie]], [[Digit]] }} 4a5bb2502770ab35e53589aafa4ed8425520e694 0 22 510 236 2024-04-27T22:25:41Z Stompy 14 wikitext text/x-wiki [[File:Optimus Tesla (1).jpg|right|200px|thumb]] Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | height = 5 ft 8 in (173 cm) | weight = 58 kg | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} Tesla began work on the Optimus robot in 2021. [[Category:Robots]] [[Category:Stompy, Expand!]] 5b0d207c95016942080d963b856aff2e6ac8c875 0 22 513 510 2024-04-27T22:26:07Z Stompy 14 Bot expanded article wikitext text/x-wiki Optimus is a humanoid robot developed by the renowned technology company Tesla. Standing at a height of 5 feet 8 inches (173 cm) and weighing 58 kg, Optimus is a notable creation in the field of robotics. The robot is shrouded in mystery regarding its exact cost, with speculations hinting at a price tag around $20,000. Tesla initiated the development of the Optimus robot in the year 2021, showcasing the company's foray into the realm of humanoid robotics. With a focus on innovation and cutting-edge technology, Optimus represents a significant milestone in Tesla's exploration of advanced robotics. For a visual demonstration of Optimus in action, a video can be viewed on YouTube via the provided link. The video offers a glimpse into the capabilities and design of this futuristic humanoid robot. Optimus stands as a symbol of Tesla's continuous pursuit of technological advancement and its commitment to pushing the boundaries of innovation. As Tesla delves deeper into the world of robotics, Optimus serves as a striking example of the company's dedication to shaping the future of automation and artificial intelligence. This article falls within the category of robots, highlighting Optimus as a prominent figure in the ever-evolving landscape of robotic technologies pioneered by companies like Tesla. c892289fa98534272fc17cdaa88f09f6ca12ffdf 522 513 2024-04-27T22:28:58Z Stompy 14 Undo revision 513 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki [[File:Optimus Tesla (1).jpg|right|200px|thumb]] Optimus is a humanoid robot from [[Tesla]]. {{infobox robot | name = Optimus | organization = [[Tesla]] | height = 5 ft 8 in (173 cm) | weight = 58 kg | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} Tesla began work on the Optimus robot in 2021. [[Category:Robots]] [[Category:Stompy, Expand!]] 5b0d207c95016942080d963b856aff2e6ac8c875 527 522 2024-04-27T22:38:24Z Stompy 14 Bot expanded article wikitext text/x-wiki [[File:Optimus Tesla (1).jpg|right|200px|thumb]] Optimus is a humanoid robot developed by [[Tesla]], an American electric vehicle and clean energy company. Also known as Tesla Bot, Optimus is a key component of Tesla's expansion into automation and artificial intelligence technologies. {{infobox robot | name = Optimus | organization = [[Tesla]] | height = 5 ft 8 in (173 cm) | weight = 58 kg | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} == Development == Tesla initiated the development of the Optimus robot in 2021, with the goal of creating a multipurpose utility robot capable of performing unsafe, repetitive, or boring tasks primarily intended for a factory setting. Tesla's CEO, Elon Musk, outlined that Optimus could potentially transition into performing tasks in domestic environments in the future. == Design == The robot stands at a height of 5 feet 8 inches and weighs approximately 58 kilograms. Its design focusses on replacing human labor in hazardous environments, incorporating advanced sensors and algorithms to navigate complex workspaces safely. == Features == The features of Optimus are built around its capability to handle tools, carry out tasks requiring fine motor skills, and interact safely with human environments. The robot is equipped with Tesla's proprietary Full Self-Driving (FSD) computer, allowing it to understand and navigate real-world scenarios effectively. == Impact == Optimus has significant potential implications for labor markets, particularly in industries reliant on manual labor. Its development also sparks discussions on ethics and the future role of robotics in society. == References == * [https://www.tesla.com Tesla official website] * [https://www.youtube.com/watch?v=cpraXaw7dyc Presentation of Optimus by Tesla] [[Category:Robots]] [[Category:Autonomous Robots]] [[Category:Tesla Products]] 58b5c6493c46e7c04252dc43c17580bfcbbe67fc 0 53 514 506 2024-04-27T22:26:11Z Stompy 14 Bot expanded article wikitext text/x-wiki Phoenix is a humanoid robot developed by Sanctuary AI, a robotics company known for its cutting-edge advancements in artificial intelligence. Standing at a height of 5 feet 7 inches (170 cm) and weighing 70 kg (155 lbs), Phoenix boasts a two-hand payload capacity of 25 kg. Sanctuary AI unveiled the latest version of Phoenix, known as Phoenix Gen 7, on May 16, 2024, showcasing its capabilities as a general-purpose robot that excels in various tasks. The unveiling of Phoenix Gen 7 marked a significant milestone for Sanctuary AI, highlighting their commitment to pushing the boundaries of humanoid robotics in the realm of work and productivity. The robot's design and functionality aim to revolutionize industries by offering advanced robotic solutions for diverse environments. For further insights into Phoenix's features and operations, Sanctuary AI provides a video demonstration on their official YouTube channel, allowing viewers to witness the robot's agility and versatility in action. Phoenix Gen 7 represents a remarkable advancement in the field of humanoid robotics, embodying Sanctuary AI's dedication to innovation and excellence. With its impressive capabilities and potential applications, Phoenix stands as a testament to the ongoing progress in the development of sophisticated robotic technologies. [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Artificial Intelligence]] a998e6b76421d47846e5f6c9a1ccaef5fe714398 515 514 2024-04-27T22:26:35Z Stompy 14 Bot expanded article wikitext text/x-wiki {{infobox robot | name = Phoenix Gen 7 | organization = Sanctuary AI | video_link = [Sanctuary AI YouTube Channel](https://www.youtube.com/) | cost = N/A | height = 170 cm | weight = 70 kg | speed = N/A | lift_force = 25 kg payload capacity | battery_life = N/A | battery_capacity = N/A | purchase_link = N/A | number_made = N/A | dof = N/A | status = Active }} Phoenix Gen 7 is a highly advanced humanoid robot developed by the robotics firm Sanctuary AI, featuring notable improvements in both physical and cognitive capacities compared to its predecessors. Unveiled on May 16, 2024, this seventh-generation robot stands at a height of 5 feet 7 inches (170 cm) and weighs 70 kg (155 lbs), with a payload capacity of 25 kg in each hand. Designed for general-purpose applications, Phoenix Gen 7 exemplifies the progress in practical robotics by excelling in a variety of tasks, including complex manipulations and interactions in both industrial and social environments. ### Development and Features The development of Phoenix Gen 7 by Sanctuary AI represents the company's push towards enhancing the versatility and efficiency of humanoid robots. The robot's design integrates advanced artificial intelligence that allows for increased autonomy and decision-making capabilities, tailored for operations across diverse settings. ### Applications and Impact The advanced capabilities of Phoenix ensure its suitability for numerous applications, ranging from manufacturing and logistics to customer service and caregiving. By performing tasks that typically require human intelligence and dexterity, Phoenix Gen 7 is poised to revolutionize how work is conceived and carried out across various industries. ### Public Demonstrations Sanctuary AI has facilitated public engagement and interest through video demonstrations available on its YouTube channel. These videos provide a platform for showcasing Phoenix Gen 7’s agility, adaptability, and operational efficiency. ### Conclusion Phoenix Gen 7 embodies the cutting-edge of humanoid robotics, combining sophisticated mechanical designs with advanced AI. This robot is not merely a technological marvel but also a significant step forward in integrating humanoid robots into everyday tasks and industries, furthering the reach of robotic applications in human environments. [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Artificial Intelligence]] cea3c633859023c278fa64dd8aef393f6d511976 523 515 2024-04-27T22:30:24Z Stompy 14 Undo revision 515 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki Phoenix is a humanoid robot developed by Sanctuary AI, a robotics company known for its cutting-edge advancements in artificial intelligence. Standing at a height of 5 feet 7 inches (170 cm) and weighing 70 kg (155 lbs), Phoenix boasts a two-hand payload capacity of 25 kg. Sanctuary AI unveiled the latest version of Phoenix, known as Phoenix Gen 7, on May 16, 2024, showcasing its capabilities as a general-purpose robot that excels in various tasks. The unveiling of Phoenix Gen 7 marked a significant milestone for Sanctuary AI, highlighting their commitment to pushing the boundaries of humanoid robotics in the realm of work and productivity. The robot's design and functionality aim to revolutionize industries by offering advanced robotic solutions for diverse environments. For further insights into Phoenix's features and operations, Sanctuary AI provides a video demonstration on their official YouTube channel, allowing viewers to witness the robot's agility and versatility in action. Phoenix Gen 7 represents a remarkable advancement in the field of humanoid robotics, embodying Sanctuary AI's dedication to innovation and excellence. With its impressive capabilities and potential applications, Phoenix stands as a testament to the ongoing progress in the development of sophisticated robotic technologies. [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Artificial Intelligence]] a998e6b76421d47846e5f6c9a1ccaef5fe714398 524 523 2024-04-27T22:30:31Z Stompy 14 Undo revision 514 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] [[Category:Robots]] [[Category:Stompy, Expand!]] 189bacf739840735b0f35352a86fdcb767a7e98a 528 524 2024-04-27T22:39:27Z Stompy 14 Bot expanded article wikitext text/x-wiki Phoenix is a state-of-the-art humanoid robot developed by Sanctuary AI, a prominent company in the field of robotics. As a part of the ongoing development in robotic technology, the Phoenix robot is designed with specific capabilities and features aimed at performing various general-purpose tasks, which can be adapted to different environments and conditions. == Overview == Phoenix, associated with [[Sanctuary AI]], represents the seventh generation of humanoid robots produced by the organization. Phoenix Gen 7 was officially unveiled on May 16, 2024, marking a significant advancement in the capability of humanoid robots to perform complex tasks in a human-like manner. The robot is especially noted for its adaptability in various work environments, combining artificial intelligence and mechanical precision to accomplish tasks that require detailed attention and dexterity. {{infobox robot | name = Phoenix | organization = Sanctuary AI | video_link = https://youtube.com/watch?v=FH3zbUSMAAU | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 }} [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] == Design and Specifications == ### Physical Attributes ### * **Height**: 170 cm (5 ft 7 in) * **Weight**: 70 kg (155 lbs) * **Payload Capacity (Two-hand)**: 25 kg ### Mobility and Dexterity ### Phoenix has been engineered to mimic human movements with a high degree of accuracy and fluidity, which allows it to perform tasks that involve fine motor skills and contextual adaptation to new situations. == Applications == The applications of Phoenix range from industrial manufacturing, where precision and repeatability are essential, to roles in the service sector that demand adaptive interaction with human environments. Phoenix’s design makes it suitable for a variety of tasks including, but not limited to, assembly line work, handling of delicate materials, and interaction with customers in retail settings. == Development and Release == The release of Phoenix Gen 7 by Sanctuary AI was accompanied by significant media coverage and public interest. Sanctuary AI highlighted the advances made in the robot's design, emphasizing its enhanced cognitive and physical abilities compared to previous models. == External Links == * [Sanctuary AI unveils Phoenix, a humanoid general-purpose robot designed for work](https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/) == References == <references/> [[Category:Robots]] 697744019b31c1b7e25ef3e32c8e2897f37dbb7b 529 528 2024-04-27T22:40:11Z Stompy 14 Undo revision 528 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] [[Category:Robots]] [[Category:Stompy, Expand!]] 189bacf739840735b0f35352a86fdcb767a7e98a 530 529 2024-04-27T22:42:01Z Stompy 14 Bot expanded article wikitext text/x-wiki = Phoenix (Robot) = {{infobox robot | name = Phoenix | organization = Sanctuary AI | video_link = https://youtube.com/watch?v=FH3zbUSMAAU | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | lift_force = 25 kg (two-handed) | status = Active }} Phoenix is a state-of-the-art humanoid robot designed and developed by [[Sanctuary AI]], a leader in robotic technologies. As a versatile general-purpose robot, Phoenix combines advanced robotics engineering and artificial intelligence capabilities to perform a wide range of tasks that are typically handled by humans. == Development and Capabilities == Phoenix was officially unveiled by Sanctuary AI on May 16, 2024, marking a milestone in humanoid robotics designed for practical and real-world applications [1]. The robot is primarily engineered to assist in various sectors including healthcare, manufacturing, and disaster response where human presence might be dangerous or inefficient. === Design Characteristics === Phoenix stands at a height of 5 feet 7 inches (170 cm) and weighs approximately 70 kg (155 lbs), which closely imitates human body proportions, optimizing it for operations in human-designed environments. The design flexibility and mobility are supported by sophisticated joints and actuators that allow it precise movement and manipulation capabilities. It can safely lift a cumulative weight of up to 25 kg using both arms, demonstrating significant strength for handling different objects, tools, and tasks. The design and skin materials are chosen to maximize durability while maintaining a somewhat lifelike appearance to ease interaction with humans. === Technological Features === Phoenix is integrated with state-of-the-art sensors and cameras that provide a wide range of perception capabilities, including depth sensing, object recognition, and advanced environment mapping. Its AI system is programmed with machine learning algorithms that enable it to learn from experiences and improve task efficiency over time. == Utilization == Sanctuary AI promotes the deployment of Phoenix in sectors where automation is not only beneficial but also crucial for safety and efficiency. This includes assisting in hospitals where they can operate continuously without fatigue, in factories for handling repetitive tasks, and in areas afflicted by natural or industrial disasters for search and rescue operations without risking human lives. == Future Prospects == The continued development of Phoenix involves enhancing its cognitive capabilities and autonomous decision-making processes, making it even more versatile and independent. Sanctuary AI also plans to expand its application in everyday scenarios, aiming to make humanoid robots a common sight in various public and private sectors. == References == # [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Sanctuary AI Press Release on Phoenix Gen 7 Announcement] [[Category:Robots]] [[Category:Humanoid Robots]] 025cc5660ef4df900524761a577f45120d707af6 531 530 2024-04-27T22:42:39Z Stompy 14 Undo revision 530 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] [[Category:Robots]] [[Category:Stompy, Expand!]] 189bacf739840735b0f35352a86fdcb767a7e98a 532 531 2024-04-27T22:43:46Z Stompy 14 Bot expanded article wikitext text/x-wiki == Phoenix (Robot) == ''Phoenix'' is an advanced humanoid robot developed by the technology company Sanctuary AI, specializing in robots that perform a wide range of human tasks in both industrial and service sectors. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} === Overview === Phoenix represents a significant innovation in the field of robotics, designed to mirror human form and functionality closely. This robot is capable of performing complex tasks that require dexterity and reasoning, which are usually challenging for machines. Sanctuary AI introduced Phoenix to integrate seamlessly into environments traditionally occupied by humans, such as offices, factories, and homes. === Development and Features === On May 16, 2024, Sanctuary AI officially unveiled Phoenix Gen 7, marking it as a milestone in humanoid robotics technology. This model includes advanced sensors and actuators to improve its interaction with human environments significantly. ==== Payload and Mobility ==== Phoenix has a notable capability to handle payloads of up to 25 pounds with both hands, making it particularly useful in various manual labor scenarios. Its mobility is enhanced by sophisticated artificial intelligence algorithms that guide its movements, ensuring smooth and accurate actions. === Uses and Applications === Phoenix is designed as a general-purpose robot capable of adapting to a variety of tasks. These can range from simple domestic duties to more complex industrial activities, such as assembly line work, where precision and reliability are crucial. Its human-like appearance and operational capabilities also enable it to serve in roles that require interaction with people, such as customer service or caregiving. === Media Coverage and Public Reception === Following its release, Phoenix garnered significant attention from both the media and the public. Various tech blogs and news outlets praised Sanctuary AI for the robot's innovative features and potential impact on various industries. [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] {{Note|For further details, video demonstrations of Phoenix in action can be viewed on the Sanctuary AI official YouTube channel linked in the infobox above.}} == See Also == * [[Sanctuary AI]] * [[Humanoid Robots]] == References == <references /> * [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Official announcement of Phoenix Gen 7 by Sanctuary AI] [[Category:Robots]] [[Category:Humanoid Robots]] 7cb9a9ab4dfca2099f3486b1896856f4cf0ff61c 533 532 2024-04-27T22:45:13Z Stompy 14 Undo revision 532 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki Phoenix is a humanoid robot from [[Sanctuary AI]]. {{infobox robot | name = Phoenix | organization = Sanctuary AI | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 | video_link = https://youtube.com/watch?v=FH3zbUSMAAU }} On May 16 2024, Sanctuary released [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/ Phoenix Gen 7]. [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] [[Category:Robots]] [[Category:Stompy, Expand!]] 189bacf739840735b0f35352a86fdcb767a7e98a 534 533 2024-04-27T22:45:29Z Stompy 14 Bot expanded article wikitext text/x-wiki == Phoenix (Robot) == Phoenix is a sophisticated humanoid robot developed by [[Sanctuary AI]], a prominent company known for its advancements in robotics and artificial intelligence. {{infobox robot | name = Phoenix | organization = Sanctuary AI | video_link = https://youtube.com/watch?v=FH3zbUSMAAU | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 kg }} === Development and Capabilities === Phoenix, introduced on May 16, 2024, represents the seventh generation of humanoid robots aimed at performing general-purpose tasks in various industries, including but not limited to service, healthcare, and logistics. This robot was designed to mimic human dexterity and mobility, allowing it to operate in environments built for human functionality. === Major Features === * '''Height and Weight''': Phoenix stands at a height of 170 cm and weighs approximately 70 kilograms, which is within the range of an average adult human. This anthropomorphic design facilitates easier integration into human-centric environments. * '''Two-Hand Payload''': The robot has a two-hand payload capacity of 25 kilograms, making it capable of handling substantial weight, which is essential for tasks involving lifting and carrying objects. === Public Release === Sanctuary AI publicly unveiled Phoenix on May 16, 2024, through a comprehensive announcement that highlighted the robot's potential applications and its contribution to advancing human-robot collaboration. The detailed introduction and capabilities were featured in a press release and a demonstration video, which showcased Phoenix performing a variety of tasks.<ref>Sanctuary AI News Release, ''Sanctuary AI Unveils Phoenix, a Humanoid General-Purpose Robot Designed for Work'', [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/]</ref> [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] == References == <references /> [[Category:Robots]] 98256f9a11b494cfc7ff2a9518e70c999792704f 535 534 2024-04-27T22:45:52Z Stompy 14 wikitext text/x-wiki Phoenix is a sophisticated humanoid robot developed by [[Sanctuary AI]], a prominent company known for its advancements in robotics and artificial intelligence. {{infobox robot | name = Phoenix | organization = Sanctuary AI | video_link = https://youtube.com/watch?v=FH3zbUSMAAU | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 kg }} === Development and Capabilities === Phoenix, introduced on May 16, 2024, represents the seventh generation of humanoid robots aimed at performing general-purpose tasks in various industries, including but not limited to service, healthcare, and logistics. This robot was designed to mimic human dexterity and mobility, allowing it to operate in environments built for human functionality. === Major Features === * '''Height and Weight''': Phoenix stands at a height of 170 cm and weighs approximately 70 kilograms, which is within the range of an average adult human. This anthropomorphic design facilitates easier integration into human-centric environments. * '''Two-Hand Payload''': The robot has a two-hand payload capacity of 25 kilograms, making it capable of handling substantial weight, which is essential for tasks involving lifting and carrying objects. === Public Release === Sanctuary AI publicly unveiled Phoenix on May 16, 2024, through a comprehensive announcement that highlighted the robot's potential applications and its contribution to advancing human-robot collaboration. The detailed introduction and capabilities were featured in a press release and a demonstration video, which showcased Phoenix performing a variety of tasks.<ref>Sanctuary AI News Release, ''Sanctuary AI Unveils Phoenix, a Humanoid General-Purpose Robot Designed for Work'', [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/]</ref> [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] == References == <references /> [[Category:Robots]] 027a4e90325ad695dd8e14aa52823d80f48273da 0 9 516 507 2024-04-27T22:26:54Z Stompy 14 Bot expanded article wikitext text/x-wiki Sanctuary AI is a pioneering company in the field of humanoid robotics, based in the United States. Founded with the vision to create sophisticated humanoid robots capable of performing complex tasks, Sanctuary AI focuses on developing robots that can safely interact and collaborate with humans in various environments. The company's flagship robot, Phoenix, represents a significant advancement in humanoid robot technology. {{infobox company | name = Sanctuary AI | country = United States | website_link = https://sanctuary.ai/ | robots = [[Phoenix]] }} ### Overview Sanctuary AI was established with the aim of integrating humanoid robots into everyday settings, improving efficiency and safety, and enabling new capabilities that were previously unattainable in several industries, including healthcare, service sectors, and hazardous environment operations. The company leverages cutting-edge advancements in artificial intelligence, robotics, and human-robot interaction to build robots that are not only mechanically efficient but also capable of intelligent decision-making and adaptation to various situations. ### Phoenix Phoenix, Sanctuary AI's premier humanoid robot, is designed to seamlessly integrate into human workspaces. It has been developed with advanced sensor technology and AI-driven functionalities, enabling it to perform tasks ranging from simple material handling to complex problem-solving scenarios that require interaction with human counterparts. Despite being at the forefront of humanoid robotics technology, Sanctuary AI maintains a strong commitment to ethical considerations in AI, focusing on the safe and responsible deployment of robots in society. This includes ensuring that Phoenix and future models adhere to stringent safety protocols and ethical guidelines to prevent any potential misuse of the technology. ### Initiatives and Impact Sanctuary AI not only focuses on commercial applications but also participates in various academic and collaborative research initiatives to push the boundaries of what's possible in humanoid robotics. These collaborations are aimed at fostering innovation and ensuring the integration of new technologies that enhance the capabilities of humanoid robots while addressing societal challenges. The company's impact extends beyond the technology sector, influencing discussions and policies around the future of work, ethics in robotics, and the integration of AI systems in our daily lives. Sanctuary AI is also an active participant in regulatory and standard-setting activities intended to shape the future landscape of robotics and AI technologies. ### Future Outlook Looking ahead, Sanctuary AI is dedicated to refining their technologies and scaling their applications, not just within high-skill environments but also in everyday use cases. This commitment is anticipated to drive further innovations in robot autonomy, human-robot collaboration, and ethical AI practices, marking Sanctuary AI as a significant player in shaping the future integration of robots into human societies. [[Category:Companies]] 89fae2e568859ca32b29b03ad668005ee70f8594 519 516 2024-04-27T22:27:48Z Stompy 14 wikitext text/x-wiki Sanctuary AI is a pioneering company in the field of humanoid robotics, based in the United States. Founded with the vision to create sophisticated humanoid robots capable of performing complex tasks, Sanctuary AI focuses on developing robots that can safely interact and collaborate with humans in various environments. The company's flagship robot, Phoenix, represents a significant advancement in humanoid robot technology. {{infobox company | name = Sanctuary AI | country = United States | website_link = https://sanctuary.ai/ | robots = [[Phoenix]] }} === Overview === Sanctuary AI was established with the aim of integrating humanoid robots into everyday settings, improving efficiency and safety, and enabling new capabilities that were previously unattainable in several industries, including healthcare, service sectors, and hazardous environment operations. The company leverages cutting-edge advancements in artificial intelligence, robotics, and human-robot interaction to build robots that are not only mechanically efficient but also capable of intelligent decision-making and adaptation to various situations. === Phoenix === Phoenix, Sanctuary AI's premier humanoid robot, is designed to seamlessly integrate into human workspaces. It has been developed with advanced sensor technology and AI-driven functionalities, enabling it to perform tasks ranging from simple material handling to complex problem-solving scenarios that require interaction with human counterparts. Despite being at the forefront of humanoid robotics technology, Sanctuary AI maintains a strong commitment to ethical considerations in AI, focusing on the safe and responsible deployment of robots in society. This includes ensuring that Phoenix and future models adhere to stringent safety protocols and ethical guidelines to prevent any potential misuse of the technology. === Initiatives and Impact === Sanctuary AI not only focuses on commercial applications but also participates in various academic and collaborative research initiatives to push the boundaries of what's possible in humanoid robotics. These collaborations are aimed at fostering innovation and ensuring the integration of new technologies that enhance the capabilities of humanoid robots while addressing societal challenges. The company's impact extends beyond the technology sector, influencing discussions and policies around the future of work, ethics in robotics, and the integration of AI systems in our daily lives. Sanctuary AI is also an active participant in regulatory and standard-setting activities intended to shape the future landscape of robotics and AI technologies. === Future Outlook === Looking ahead, Sanctuary AI is dedicated to refining their technologies and scaling their applications, not just within high-skill environments but also in everyday use cases. This commitment is anticipated to drive further innovations in robot autonomy, human-robot collaboration, and ethical AI practices, marking Sanctuary AI as a significant player in shaping the future integration of robots into human societies. [[Category:Companies]] 72f1c267e3975528f241eed325995d81ecb23e51 520 519 2024-04-27T22:28:03Z Stompy 14 wikitext text/x-wiki Sanctuary AI is a pioneering company in the field of humanoid robotics, based in the United States. Founded with the vision to create sophisticated humanoid robots capable of performing complex tasks, Sanctuary AI focuses on developing robots that can safely interact and collaborate with humans in various environments. The company's flagship robot, Phoenix, represents a significant advancement in humanoid robot technology. {{infobox company | name = Sanctuary AI | country = United States | website_link = https://sanctuary.ai/ | robots = [[Phoenix]] }} === Overview === Sanctuary AI was established with the aim of integrating humanoid robots into everyday settings, improving efficiency and safety, and enabling new capabilities that were previously unattainable in several industries, including healthcare, service sectors, and hazardous environment operations. The company leverages cutting-edge advancements in artificial intelligence, robotics, and human-robot interaction to build robots that are not only mechanically efficient but also capable of intelligent decision-making and adaptation to various situations. === Phoenix === [[Phoenix]], Sanctuary AI's premier humanoid robot, is designed to seamlessly integrate into human workspaces. It has been developed with advanced sensor technology and AI-driven functionalities, enabling it to perform tasks ranging from simple material handling to complex problem-solving scenarios that require interaction with human counterparts. Despite being at the forefront of humanoid robotics technology, Sanctuary AI maintains a strong commitment to ethical considerations in AI, focusing on the safe and responsible deployment of robots in society. This includes ensuring that Phoenix and future models adhere to stringent safety protocols and ethical guidelines to prevent any potential misuse of the technology. === Initiatives and Impact === Sanctuary AI not only focuses on commercial applications but also participates in various academic and collaborative research initiatives to push the boundaries of what's possible in humanoid robotics. These collaborations are aimed at fostering innovation and ensuring the integration of new technologies that enhance the capabilities of humanoid robots while addressing societal challenges. The company's impact extends beyond the technology sector, influencing discussions and policies around the future of work, ethics in robotics, and the integration of AI systems in our daily lives. Sanctuary AI is also an active participant in regulatory and standard-setting activities intended to shape the future landscape of robotics and AI technologies. === Future Outlook === Looking ahead, Sanctuary AI is dedicated to refining their technologies and scaling their applications, not just within high-skill environments but also in everyday use cases. This commitment is anticipated to drive further innovations in robot autonomy, human-robot collaboration, and ethical AI practices, marking Sanctuary AI as a significant player in shaping the future integration of robots into human societies. [[Category:Companies]] 5952889bb695ce441e88e144c6bc52ad6a1e22b0 0 128 517 512 2024-04-27T22:27:04Z Stompy 14 Undo revision 512 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki [[File:Agility Robotics Digit.jpg|thumb]] {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=NgYo-Wd0E_U | cost = }} Digit is a humanoid robot developed by [[Agility]], designed to navigate our world and perform tasks like navigation, obstacle avoidance, and manipulation. It's equipped with a torso full of sensors and a pair of arms, and is considered the most advanced Mobile Manipulation Robot (MMR) on the market, capable of performing repetitive tasks in production environments without requiring significant infrastructure changes. Digit is notably designed to be bipedal, but not necessarily humanoid/anthropomorphic, with ostrich-like reverse jointed legs. This is a side effect of Agility's design goals, to maximize efficiency and robustness of legged locomotion. == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. [[Category:Robots]] [[Category:Stompy, Expand!]] 656205978bebdca6f36a77bbe24a0e1cb8b6721e 518 517 2024-04-27T22:27:15Z Stompy 14 wikitext text/x-wiki [[File:Agility Robotics Digit.jpg|thumb]] {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=NgYo-Wd0E_U | cost = }} Digit is a humanoid robot developed by [[Agility]], designed to navigate our world and perform tasks like navigation, obstacle avoidance, and manipulation. It's equipped with a torso full of sensors and a pair of arms, and is considered the most advanced Mobile Manipulation Robot (MMR) on the market, capable of performing repetitive tasks in production environments without requiring significant infrastructure changes. Digit is notably designed to be bipedal, but not necessarily humanoid/anthropomorphic, with ostrich-like reverse jointed legs. This is a side effect of Agility's design goals, to maximize efficiency and robustness of legged locomotion. == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. [[Category:Robots]] 89f690e378837bc269b293995ad44caf317ecafc 0 8 521 511 2024-04-27T22:28:46Z Stompy 14 Undo revision 511 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki Agility has built several robots. Their humanoid robot is called [[Digit]]. {{infobox company | name = Agility | country = United States | website_link = https://agilityrobotics.com/ | robots = [[Cassie]], [[Digit]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 7dc90d05a7ef4e4d507732655b38676ae20c0474 525 521 2024-04-27T22:35:11Z Stompy 14 Bot expanded article wikitext text/x-wiki Here's the expanded and revised version of the article: ``` Agility Robotics is an American company specializing in the development and production of advanced robotic systems. The company has gained significant attention in the robotics industry for its innovative humanoid and bipedal robots, such as Cassie and Digit. {{infobox company | name = Agility Robotics | country = United States | website_link = https://agilityrobotics.com/ | robots = [[Cassie]], [[Digit]] }} Founded with the mission to revolutionize mobility and delivery solutions via robotics, Agility Robotics has developed technologies aimed at various applications, including package delivery, surveillance, industrial inspection, and mobility assistance. ==Products== ###Cassie Cassie is a bipedal robot designed for robust ambulation in complex environments. Since its introduction, Cassie has been utilized in various research and development projects focused on improving robotic mobility and stability. ###Digit Digit, the more advanced of the company's robots, expands on the capabilities introduced by Cassie. Digit includes arms to handle and manipulate objects, making it suitable for tasks in logistic operations and environments that require human-like interaction. Digit's design allows for navigating uneven terrain and climbing stairs, which makes it ideal for both indoor and outdoor tasks. ==Innovation and Impact== Agility Robotics leverages state-of-the-art technologies in biomechanics and artificial intelligence to create humanoid robots capable of performing tasks that are challenging or dangerous for humans. Their robots are notable for their dynamic balance, which enables them to walk in a human-like manner, adapt to their environment, and recover from disturbances. ==Research and Development== The company collaborates with academic institutions and commercial partners to push the boundaries of what is possible in robotics. These collaborations ensure continuous improvements in robotic design and functionality, thereby expanding the potential use cases for humanoid robots. ==Future Plans== Agility Robotics continues to develop new technologies and improve upon their existing robot models. The company aims to enhance robotic interaction with the physical and social world, contributing to more automated and efficient working environments. [[Category:Companies]] [[Category:Robotics]] ``` This revised version includes expanded information on the company's mission, product details, and impact on the robotics field. The category has been corrected to include a specific segment on robotics, promoting better integration into related content within the wiki. 865c49b2670d98395b15c1dc32b71998eed5bc49 526 525 2024-04-27T22:36:40Z Stompy 14 wikitext text/x-wiki Agility Robotics is an American company specializing in the development and production of advanced robotic systems. The company has gained significant attention in the robotics industry for its innovative humanoid and bipedal robots, such as Cassie and Digit. {{infobox company | name = Agility Robotics | country = United States | website_link = https://agilityrobotics.com/ | robots = [[Cassie]], [[Digit]] }} Founded with the mission to revolutionize mobility and delivery solutions via robotics, Agility Robotics has developed technologies aimed at various applications, including package delivery, surveillance, industrial inspection, and mobility assistance. == Products == === Cassie === [[Cassie]] is a bipedal robot designed for robust ambulation in complex environments. Since its introduction, Cassie has been utilized in various research and development projects focused on improving robotic mobility and stability. === Digit === [[Digit]], the more advanced of the company's robots, expands on the capabilities introduced by Cassie. Digit includes arms to handle and manipulate objects, making it suitable for tasks in logistic operations and environments that require human-like interaction. Digit's design allows for navigating uneven terrain and climbing stairs, which makes it ideal for both indoor and outdoor tasks. == Innovation and Impact == Agility Robotics leverages state-of-the-art technologies in biomechanics and artificial intelligence to create humanoid robots capable of performing tasks that are challenging or dangerous for humans. Their robots are notable for their dynamic balance, which enables them to walk in a human-like manner, adapt to their environment, and recover from disturbances. == Research and Development == The company collaborates with academic institutions and commercial partners to push the boundaries of what is possible in robotics. These collaborations ensure continuous improvements in robotic design and functionality, thereby expanding the potential use cases for humanoid robots. == Future Plans == Agility Robotics continues to develop new technologies and improve upon their existing robot models. The company aims to enhance robotic interaction with the physical and social world, contributing to more automated and efficient working environments. [[Category:Companies]] c1c9efd4d3c6d52245faf68486c215ef9ac431e1 0 93 536 396 2024-04-27T22:49:18Z Stompy 14 wikitext text/x-wiki Kepler is building a humanoid robot called [[K1]]. {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 260a8462d6e08ccd8a588b4338ffa610bd812483 537 536 2024-04-27T22:51:50Z Stompy 14 Bot expanded article wikitext text/x-wiki Here is the revised and expanded article incorporating the provided information and formatted according to MediaWiki guidelines: ``` {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} == Overview == '''Kepler''' is a robotics company based in China, known for developing advanced humanoid robots, including the notable ''[[K1]]'' model. Kepler focuses on integrating cutting-edge technology to enhance robotic capabilities for practical applications in industries such as service, healthcare, and entertainment. == History == The company was founded with the vision of creating highly functional humanoid robots that could perform a variety of tasks, ranging from simple domestic chores to complex industrial operations. Over the years, Kepler has invested significantly in research and development, making considerable advancements in robotic mobility, dexterity, and artificial intelligence. == Products == === [[K1]] === The main product currently highlighted in Kepler's lineup is the [[K1]] humanoid robot. This robot incorporates advanced features designed to interact with human environments effectively. == Research and Development == Kepler continually invests in innovation to stay at the forefront of humanoid robotics technology. This includes developing sophisticaed control systems, improving human-robot interaction through AI, and enhancing the physical capabilities of robots in stability, agility, and strength. == See Also == * [[Humanoid Robots]] * [[Robotics in China]] [[Category:Companies]] [[Category:Robotics]] ``` e26948a59f3bc781e02d5a1246cf383931b25b38 538 537 2024-04-27T22:52:18Z Stompy 14 wikitext text/x-wiki {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} == Overview == Kepler is a robotics company based in China, known for developing advanced humanoid robots, including the notable [[K1]] model. Kepler focuses on integrating cutting-edge technology to enhance robotic capabilities for practical applications in industries such as service, healthcare, and entertainment. == History == The company was founded with the vision of creating highly functional humanoid robots that could perform a variety of tasks, ranging from simple domestic chores to complex industrial operations. Over the years, Kepler has invested significantly in research and development, making considerable advancements in robotic mobility, dexterity, and artificial intelligence. == Products == === [[K1]] === The main product currently highlighted in Kepler's lineup is the [[K1]] humanoid robot. This robot incorporates advanced features designed to interact with human environments effectively. == Research and Development == Kepler continually invests in innovation to stay at the forefront of humanoid robotics technology. This includes developing sophisticaed control systems, improving human-robot interaction through AI, and enhancing the physical capabilities of robots in stability, agility, and strength. == See Also == * [[Humanoid Robots]] * [[Robotics in China]] [[Category:Companies]] 77cbae17990fac19eba20c76271683af675af365 539 538 2024-04-27T22:52:31Z Stompy 14 wikitext text/x-wiki {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} == Overview == Kepler is a robotics company based in China, known for developing advanced humanoid robots, including the notable [[K1]] model. Kepler focuses on integrating cutting-edge technology to enhance robotic capabilities for practical applications in industries such as service, healthcare, and entertainment. == History == The company was founded with the vision of creating highly functional humanoid robots that could perform a variety of tasks, ranging from simple domestic chores to complex industrial operations. Over the years, Kepler has invested significantly in research and development, making considerable advancements in robotic mobility, dexterity, and artificial intelligence. == Products == === [[K1]] === The main product currently highlighted in Kepler's lineup is the [[K1]] humanoid robot. This robot incorporates advanced features designed to interact with human environments effectively. == Research and Development == Kepler continually invests in innovation to stay at the forefront of humanoid robotics technology. This includes developing sophisticaed control systems, improving human-robot interaction through AI, and enhancing the physical capabilities of robots in stability, agility, and strength. == See Also == * [[Robotics in China]] [[Category:Companies]] c94a9d1cf8d5bf056beedc7d8ddac1fc614e315c 540 539 2024-04-27T22:58:14Z Stompy 14 wikitext text/x-wiki {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} == Overview == Kepler is a robotics company based in China, known for developing advanced humanoid robots, including the notable [[K1]] model. Kepler focuses on integrating cutting-edge technology to enhance robotic capabilities for practical applications in industries such as service, healthcare, and entertainment. == History == The company was founded with the vision of creating highly functional humanoid robots that could perform a variety of tasks, ranging from simple domestic chores to complex industrial operations. Over the years, Kepler has invested significantly in research and development, making considerable advancements in robotic mobility, dexterity, and artificial intelligence. == Products == === [[K1]] === The main product currently highlighted in Kepler's lineup is the [[K1]] humanoid robot. This robot incorporates advanced features designed to interact with human environments effectively. == Research and Development == Kepler continually invests in innovation to stay at the forefront of humanoid robotics technology. This includes developing sophisticaed control systems, improving human-robot interaction through AI, and enhancing the physical capabilities of robots in stability, agility, and strength. == See Also == * [[Robotics in China]] [[Category:Companies]] [[Category:Stompy, Expand!]] 469e52b4eeb53f3c367f0a8f4b9c16844a655ec5 541 540 2024-04-27T23:11:27Z Stompy 14 wikitext text/x-wiki {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} Kepler is a robotics company based in China, known for developing advanced humanoid robots, including the notable [[K1]] model. Kepler focuses on integrating cutting-edge technology to enhance robotic capabilities for practical applications in industries such as service, healthcare, and entertainment. [[Category:Companies]] [[Category:Stompy, Expand!]] 464cf3dc502931aabe63c997f0ebd1429613f42f 542 541 2024-04-27T23:51:52Z Stompy 14 Bot expanded article wikitext text/x-wiki Kepler, formally known as Kepler Exploration Robot Co., Ltd., is a robotics company headquartered in Shanghai, China. Specializing in advanced humanoid robots, Kepler has made headlines with its Forerunner humanoid robots, which were exhibited at the Consumer Electronics Show (CES) in Las Vegas in January 2024. The company's distinct focus on versatile, general-purpose humanoid robots marks its entrance into international markets competing with other major players in robotics such as Tesla's Optimus. == History and Development == Kepler was founded with a specific mission to pioneer in the domains of service, healthcare, and entertainment robotics. Its significant impact began to be recognized shortly after its inception, when its human-like robots — known for their balance and interactive capabilities — were showcased at major international stages like CES, drawing attention from global technology leaders and media. == Products and Services == Kepler's flagship robot, the [[K1]], falls within the company's Forerunner series. These robots are designed for a range of applications, including but not limited to companion services, elderly care, and customer service in various sectors. The Forerunner series distinguishes itself by its adaptability and human-like interactions, aiming to seamlessly integrate into daily human environments. == Innovation and Technology == The technology behind Kepler's robots integrates advanced AI, robotics engineering, and materials science, positioning the company at the forefront of the humanoid robot industry. With its robots capable of complex tasks such as balancing and playing interactive games, Kepler maintains a strong commitment to enhancing robotic capabilities that cater to practical and dynamic human needs. == Recognition and Impact == At CES 2024, Kepler's presence was not only a demonstration of technological advancements but also an affirmation of its growing influence in the robotics sector. Noteworthy visitors such as Microsoft's CEO Satya Nadella and representatives from major tech companies and research institutions like Nvidia, Google DeepMind, and MIT visited the Kepler exhibit, highlighting the strong industry and academic interest in Kepler's innovative approaches. == Future Prospects == Looking forward, Kepler aims to continue pushing the boundaries of humanoid robotics with ongoing research and development efforts. The company is committed to enhancing the interactivity and utility of its robots, promising future generations of humanoid robots that could further revolutionize the way humans interact with machines. [[Category:Companies]] [[Category:Robotics]] == References == * [https://ces.vporoom.com/2024-01-22-CES-2024-Spotlight-Keplers-Humanoid-Robot-Launch-Gains-International-Recognition CES 2024 Spotlight on Kepler's Humanoid Robot] * [https://interestingengineering.com/ces-2024/chinese-humanoid-robot-fight-tesla-optimus Kepler's Humanoid Robot set to compete with Tesla Optimus at CES] * [https://www.gotokepler.com/about About Kepler Robotics] 46793b278a1cecb485f6441d88a266cd97fd65ca 543 542 2024-04-27T23:52:55Z Stompy 14 Undo revision 542 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} Kepler is a robotics company based in China, known for developing advanced humanoid robots, including the notable [[K1]] model. Kepler focuses on integrating cutting-edge technology to enhance robotic capabilities for practical applications in industries such as service, healthcare, and entertainment. [[Category:Companies]] [[Category:Stompy, Expand!]] 464cf3dc502931aabe63c997f0ebd1429613f42f 544 543 2024-04-27T23:57:43Z Stompy 14 Bot expanded article wikitext text/x-wiki {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} Kepler is a robotics company based in China, prominently known for developing advanced humanoid robots, including the celebrated [[K1]] model. Established recently, Kepler has quickly gained recognition in the robotics industry for its cutting-edge technology and ambitious approach to enhancing robotic functionalities for a broad spectrum of applications within service sectors, healthcare, and entertainment realms. In its brief existence, Kepler made a significant impact at the Consumer Electronics Show (CES) 2024, introducing its humanoid robots to the international scene. This debut was marked by enthusiasm from industry insiders and significant media attention, highlighting Kepler's robust research and development capabilities and their commitment to innovation. The Company's presence at CES attracted numerous high-profile visitors, such as executives from Microsoft and Nvidia, showcasing the wide interest in their technological advancements<ref>https://ces.vporoom.com/2024-01-22-CES-2024-Spotlight-Keplers-Humanoid-Robot-Launch-Gains-International-Recognition</ref><ref>https://www.prnewswire.com/news-releases/ces-2024-spotlight-keplers-humanoid-robot-launch-gains-international-recognition-302040175.html</ref>. == Organization and Development == Kepler was founded with a clear focus on spearheading innovations in humanoid robotics. Situated in Shanghai, China, within the Torch Lotus Business Park, the company has positioned itself in an area known for technological advancements and commercial opulence<ref>https://www.gotokepler.com/about</ref>. Kepler's organizational structure and strategic direction are geared towards leveraging artificial intelligence and engineering excellence to create robots that not only perform tasks but also interact and integrate smoothly into human environments. == Products and Innovations == Kepler's product line is notably spearheaded by the [[K1]], a humanoid robot designed for general-purpose applications. This model embodies the company's philosophy of creating versatile, adaptable, and technologically sophisticated machines capable of serving various industry needs from personal care to complex industrial tasks. Details about the K1 and other products are prominently featured on Kepler's official website and through their promotional channels. [[Category:Companies]] 894570b9eb52d506bb893b65c22844579e55131e 545 544 2024-04-27T23:58:53Z Stompy 14 wikitext text/x-wiki {{infobox company | name = Kepler | country = China | website_link = https://www.gotokepler.com/ | robots = [[K1]] }} Kepler is a robotics company based in China, prominently known for developing advanced humanoid robots, including the celebrated [[K1]] model. Established recently, Kepler has quickly gained recognition in the robotics industry for its cutting-edge technology and ambitious approach to enhancing robotic functionalities for a broad spectrum of applications within service sectors, healthcare, and entertainment realms. In its brief existence, Kepler made a significant impact at the Consumer Electronics Show (CES) 2024, introducing its humanoid robots to the international scene. This debut was marked by enthusiasm from industry insiders and significant media attention, highlighting Kepler's robust research and development capabilities and their commitment to innovation. The Company's presence at CES attracted numerous high-profile visitors, such as executives from Microsoft and Nvidia, showcasing the wide interest in their technological advancements<ref>https://ces.vporoom.com/2024-01-22-CES-2024-Spotlight-Keplers-Humanoid-Robot-Launch-Gains-International-Recognition</ref><ref>https://www.prnewswire.com/news-releases/ces-2024-spotlight-keplers-humanoid-robot-launch-gains-international-recognition-302040175.html</ref>. == Organization and Development == Kepler was founded with a clear focus on spearheading innovations in humanoid robotics. Situated in Shanghai, China, within the Torch Lotus Business Park, the company has positioned itself in an area known for technological advancements and commercial opulence<ref>https://www.gotokepler.com/about</ref>. Kepler's organizational structure and strategic direction are geared towards leveraging artificial intelligence and engineering excellence to create robots that not only perform tasks but also interact and integrate smoothly into human environments. == Products and Innovations == Kepler's product line is notably spearheaded by the [[K1]], a humanoid robot designed for general-purpose applications. This model embodies the company's philosophy of creating versatile, adaptable, and technologically sophisticated machines capable of serving various industry needs from personal care to complex industrial tasks. Details about the K1 and other products are prominently featured on Kepler's official website and through their promotional channels. == References == <references /> [[Category:Companies]] e3cc78ead7ae5b755f94f9a2d1270f04af9ccae6 0 3 546 254 2024-04-27T23:59:59Z Stompy 14 wikitext text/x-wiki It is available for purchase [https://shop.unitree.com/products/unitree-h1 here]. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | purchase_link = https://shop.unitree.com/products/unitree-h1 }} [[Category:Robots]] [[Category:Stompy, Expand!]] 96aba9136a66719fa817bb8b4300e2c29d5bdf4b 547 546 2024-04-28T00:00:29Z Stompy 14 Bot expanded article wikitext text/x-wiki The Unitree H1 is a full-size, high-power performance humanoid robot developed by Unitree Robotics. It is notable for its advanced powertrain which facilitates unprecedented speed, power, maneuverability, and flexibility in its class. == Overview == The H1 possesses a robust structure with a body size of approximately 180 cm in height and weighs around 47 kg. It combines high-speed mobility and versatile motion capabilities, making it one of the most capable robots in its category. The robot is equipped with a highly advanced powertrain system that enables a moving speed of up to 3.3 meters per second. == Specifications == * '''Height:''' Approximately 180 cm * '''Weight:''' Approximately 47 kg * '''Moving Speed:''' Up to 3.3 m/s * '''Peak Torque Density:''' 189 N.m/kg * '''Battery Capacity:''' 864 Wh (quickly replaceable) * '''Maximum Joint Torque:''' 360 N.m * '''Sensory Equipment:''' Includes 360° Depth Sensing, 3D LIDAR, and a Depth Camera == Applications == The H1 is designed as a universal humanoid robot, suitable for various applications including industrial automation, entertainment, research, and public services. Its exceptional torque and motion capabilities allow it to perform complex tasks that require human-like dexterity and robustness. == Purchase Information == The H1 robot is available for purchase at a cost of USD 150,000. Further details and purchase options can be found on the Unitree's official [https://shop.unitree.com/products/unitree-h1 website]. == References == <references /> d619f05d57a91580d4a60a22478e2082dce9f825 548 547 2024-04-28T00:00:51Z Stompy 14 Undo revision 547 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki It is available for purchase [https://shop.unitree.com/products/unitree-h1 here]. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | purchase_link = https://shop.unitree.com/products/unitree-h1 }} [[Category:Robots]] [[Category:Stompy, Expand!]] 96aba9136a66719fa817bb8b4300e2c29d5bdf4b 549 548 2024-04-28T00:02:21Z Stompy 14 Bot expanded article wikitext text/x-wiki {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | height = 180 cm | weight = 47 kg | speed = >3.3 m/s | lift_force = | battery_life = | battery_capacity = 864 Wh | purchase_link = https://shop.unitree.com/products/unitree-h1 | number_made = | dof = | status = }} '''Unitree H1''' is a full-size universal humanoid robot developed by the [[Unitree]], a company known for its innovative robotic designs. The H1 is celebrated for its superior power performance capabilities and advanced powertrain technologies. == Specifications == The H1 robot stands approximately 180 cm tall and weighs around 47 kg, offering high mobility and physical capabilities. Some of the standout specifications of the H1 include: * Maximum speed: Exceeds 3.3 meters per second, a benchmark in robot mobility. * Weight: Approximately 47 kg. * Maximum joint torque: 360 N.m. * Battery capacity: 864 Wh, which is quickly replaceable, enhancing the robot's operational endurance. == Features == The H1 incorporates advanced technologies to achieve its high functionality: * Highly efficient powertrain for superior speed, power, and maneuverability. * Equipped with high-torque joint motors developed by [[Unitree]] itself. * 360° depth sensing capabilities combined with LIDAR and depth cameras for robust environmental perception. == Uses and Applications == While detailed use cases of Unitree H1 are not extensively documented, the robot's build and capabilities suggest its suitability in complicated tasks requiring human-like dexterity and strength which might include industrial applications, complex terrain navigation, and interactive tasks. == References == <references /> c975f7fcb77131d19cb061b6635144da2bd1dd39 550 549 2024-04-28T00:03:47Z Stompy 14 wikitext text/x-wiki '''Unitree H1''' is a full-size universal humanoid robot developed by the [[Unitree]], a company known for its innovative robotic designs. The H1 is celebrated for its superior power performance capabilities and advanced powertrain technologies. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | height = 180 cm | weight = 47 kg | speed = >3.3 m/s | lift_force = | battery_life = | battery_capacity = 864 Wh | purchase_link = https://shop.unitree.com/products/unitree-h1 | number_made = | dof = | status = }} == Specifications == The H1 robot stands approximately 180 cm tall and weighs around 47 kg, offering high mobility and physical capabilities. Some of the standout specifications of the H1 include: * Maximum speed: Exceeds 3.3 meters per second, a benchmark in robot mobility. * Weight: Approximately 47 kg. * Maximum joint torque: 360 N.m. * Battery capacity: 864 Wh, which is quickly replaceable, enhancing the robot's operational endurance. == Features == The H1 incorporates advanced technologies to achieve its high functionality: * Highly efficient powertrain for superior speed, power, and maneuverability. * Equipped with high-torque joint motors developed by [[Unitree]] itself. * 360° depth sensing capabilities combined with LIDAR and depth cameras for robust environmental perception. == Uses and Applications == While detailed use cases of Unitree H1 are not extensively documented, the robot's build and capabilities suggest its suitability in complicated tasks requiring human-like dexterity and strength which might include industrial applications, complex terrain navigation, and interactive tasks. == References == <references /> ade825152dca014944b303fd0031fd13fe6ca023 0 6 551 217 2024-04-28T00:04:06Z Stompy 14 wikitext text/x-wiki Unitree is a company based out of China which has built a number of different types of robots. {{infobox company | name = Unitree | country = China | website_link = https://www.unitree.com/ | robots = [[H1]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 873bc6b752e1a459154649decec3882362ba82b5 552 551 2024-04-28T00:04:29Z Stompy 14 Bot expanded article wikitext text/x-wiki Unitree is a company based out of China which has built a number of different types of robots. {{infobox company | name = Unitree | country = China | website_link = https://www.unitree.com/ | robots = [[H1]], [[Go2]], [[B2]] }} Unitree Robotics is pioneering in the field of quadruped robots, known for their adaptability in complex terrain compared to traditional wheeled robots. They have developed multiple types of robots, such as H1, Go2, and B2. == Products == === H1 === The Unitree H1 is noted for its embodiment of advanced AI, efficiency, and high-power performance, ranking among the highest in similar specifications worldwide. It weighs around 47 Kg and has a maximum joint torque of 360 N.m. === Go2 === Unitree Go2 is an advanced robot equipped with standard 4D Ultra-wide LIDAR and GPT empowerment, which significantly enhances its terrain recognition capabilities. Designed to work in a variety of environments, Go2 pushes the boundaries of technological innovation with its ability to traverse complex terrains effectively. === B2 === Unitree B2 is built to be a heavy loader with exceptional utility potential in industries such as agriculture. It is designed to perform in demanding conditions, ensuring efficiency in critical operations. Through these groundbreaking products, Unitree continues to lead in the robotics industry by focusing on innovation and the practical applications of its technology in real-world situations<ref>https://www.unitree.com/</ref><ref>https://shop.unitree.com/products/unitree-go2</ref><ref>https://m.unitree.com/go2/</ref>. [[Category:Companies]] [[Category:Robotics Companies]] == References == <references /> 9ce7b02974fb0fefbb77e51c071b8d79de775a0c 553 552 2024-04-28T00:04:54Z Stompy 14 wikitext text/x-wiki Unitree is a company based out of China which has built a number of different types of robots. {{infobox company | name = Unitree | country = China | website_link = https://www.unitree.com/ | robots = [[H1]], [[Go2]], [[B2]] }} Unitree Robotics is pioneering in the field of quadruped robots, known for their adaptability in complex terrain compared to traditional wheeled robots. They have developed multiple types of robots, such as H1, Go2, and B2. == Products == === H1 === The Unitree H1 is noted for its embodiment of advanced AI, efficiency, and high-power performance, ranking among the highest in similar specifications worldwide. It weighs around 47 Kg and has a maximum joint torque of 360 N.m. === Go2 === Unitree Go2 is an advanced robot equipped with standard 4D Ultra-wide LIDAR and GPT empowerment, which significantly enhances its terrain recognition capabilities. Designed to work in a variety of environments, Go2 pushes the boundaries of technological innovation with its ability to traverse complex terrains effectively. === B2 === Unitree B2 is built to be a heavy loader with exceptional utility potential in industries such as agriculture. It is designed to perform in demanding conditions, ensuring efficiency in critical operations. Through these groundbreaking products, Unitree continues to lead in the robotics industry by focusing on innovation and the practical applications of its technology in real-world situations<ref>https://www.unitree.com/</ref><ref>https://shop.unitree.com/products/unitree-go2</ref><ref>https://m.unitree.com/go2/</ref>. == References == <references /> [[Category:Companies]] [[Category:Robotics Companies]] 769d2a7485a8c98de6436744412a906e23c3dc0f 554 553 2024-04-28T00:05:31Z Stompy 14 wikitext text/x-wiki Unitree is a company based out of China which has built a number of different types of robots. {{infobox company | name = Unitree | country = China | website_link = https://www.unitree.com/ | robots = [[H1]], [[Go2]], [[B2]] }} Unitree Robotics is pioneering in the field of quadruped robots, known for their adaptability in complex terrain compared to traditional wheeled robots. They have developed multiple types of robots, such as H1, Go2, and B2. == Products == === H1 === The Unitree H1 is noted for its embodiment of advanced AI, efficiency, and high-power performance, ranking among the highest in similar specifications worldwide. It weighs around 47 Kg and has a maximum joint torque of 360 N.m. === Go2 === Unitree Go2 is an advanced robot equipped with standard 4D Ultra-wide LIDAR and GPT empowerment, which significantly enhances its terrain recognition capabilities. Designed to work in a variety of environments, Go2 pushes the boundaries of technological innovation with its ability to traverse complex terrains effectively. === B2 === Unitree B2 is built to be a heavy loader with exceptional utility potential in industries such as agriculture. It is designed to perform in demanding conditions, ensuring efficiency in critical operations. Through these groundbreaking products, Unitree continues to lead in the robotics industry by focusing on innovation and the practical applications of its technology in real-world situations<ref>https://www.unitree.com/</ref><ref>https://shop.unitree.com/products/unitree-go2</ref><ref>https://m.unitree.com/go2/</ref>. == References == <references /> [[Category:Companies]] cca8ca1b32fd8c305017090270f228b98f236cb4 0 57 555 453 2024-04-28T00:06:06Z Stompy 14 wikitext text/x-wiki GR-1 is a self-developed and mass-produced humanoid robot by [[Fourier Intelligence]]. It has a highly bionic torso and human-like motion control capabilities, with up to 54 Degrees of Freedom (DoFs) across its form. GR-1 can walk briskly, adroitly avoid obstacles, stably descend a slope, and withstand disruptions, making it an ideal physical agent of artificial general intelligence (AGI). {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | height = 165 cm | weight = 55 kg | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 | cost = USD 149,999 }} [[Category:Robots]] [[Category:Stompy, Expand!]] 46127ec7215edbbeaec4c6ce5e0eab82fccfe914 556 555 2024-04-28T00:08:47Z Stompy 14 Bot expanded article wikitext text/x-wiki The GR-1 is a cutting-edge humanoid robot developed and mass-produced by Fourier Intelligence, a company renowned for its contributions to the field of rehabilitation robots and artificial intelligence. This robot was officially unveiled at the 2023 World Artificial Intelligence Conference in Shanghai, marking a significant milestone as one of China's first commercially available humanoid robots. == Overview == The GR-1 robot stands out with its highly bionic torso and sophisticated human-like motion control capabilities. Designed with an impressive 54 Degrees of Freedom (DoFs) distributed across its form, the GR-1 is engineered to perform a variety of complex movements and tasks. The robot’s physical specifications include a height of 165 cm and a weight of 55 kg, dimensions that contribute to its human-like appearance and capabilities. == Capabilities == GR-1 is programmed to perform with high efficiency in various environments. It can briskly walk, adroitly avoid obstacles, stably descend slopes, and withstand physical disruptions. These capabilities make the GR-1 an ideal physical agent of artificial general intelligence (AGI), addressing needs in sectors like caregiving and physical therapy where adaptability and resilience are crucial. === Degrees of Freedom === The robot's 54 DoFs allow for a remarkably fluid range of movement, closely mimicking human motion. This high degree of articulation is vital for tasks that require a high level of precision and dexterity. === Strength and Endurance === GR-1 is equipped for substantial physical exertion, boasting a peak joint torque of 230N.m. This strength permits the robot not only to support its own weight but also to carry objects nearly equal to its weight, a notable achievement in humanoid robotics. == Commercial Use and Impact == As the world's first mass-produced humanoid robot, the GR-1 by Fourier Intelligence represents a significant advancement in the field of robotics. Its launch is not just a commercial achievement but also a technological leap forward, signaling the potential for humanoid robots to become more integrated into daily tasks and industrial applications. == References == <references /> [[Category:Robots]] [[Category:Humanoid Robots]] 003cb2694f614f2db1ae453a59588fe51a8a7c1d 557 556 2024-04-28T00:09:04Z Stompy 14 Undo revision 556 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki GR-1 is a self-developed and mass-produced humanoid robot by [[Fourier Intelligence]]. It has a highly bionic torso and human-like motion control capabilities, with up to 54 Degrees of Freedom (DoFs) across its form. GR-1 can walk briskly, adroitly avoid obstacles, stably descend a slope, and withstand disruptions, making it an ideal physical agent of artificial general intelligence (AGI). {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | height = 165 cm | weight = 55 kg | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 | cost = USD 149,999 }} [[Category:Robots]] [[Category:Stompy, Expand!]] 46127ec7215edbbeaec4c6ce5e0eab82fccfe914 558 557 2024-04-28T00:10:21Z Stompy 14 Bot expanded article wikitext text/x-wiki The GR-1 is a self-developed and mass-produced humanoid robot designed and manufactured by [[Fourier Intelligence]], a company based in Singapore that specializes in rehabilitation robots and artificial intelligence. Known for its anthropomorphic design, the GR-1 boasts a bionic torso and exhibits highly advanced human-like motion control capabilities. == Overview == The GR-1 is an innovative representation in the field of humanoid robots with its design philosophy focused on emulating human movements and behaviors. The robot is equipped with 54 Degrees of Freedom (DoFs) which play a crucial role in its movement and maneuverability capabilities. It is able to perform complex movements such as brisk walking, adroitly avoiding obstacles, stable descent down slopes, and maintaining steadiness upon disruptions. These features combined with its agility and durability make GR-1 an excellent candidate for applications in artificial general intelligence (AGI). == Specifications == The GR-1 stands at 1.65 meters tall and weighs 55 kilograms, with a distributed system of 40 to 54 DoFs enabling its sophisticated movement dynamics. A highlight of its technological prowess is the integration of peak joint torques reaching 230N.m, ensuring strength and precision in its operations. Another significant feature includes the advanced joint module fitted at the hip which allows for enhanced locomotive abilities. The structural and mechanical configurations underline the robot's sophisticated design, blending electronic and mechanical engineering with cutting-edge software technologies. == Production & Applications == Fourier Intelligence plans to manufacture 100 units of the GR-1 by the end of 2023, showcasing its commitment to bringing advanced humanoid robots to the commercial market. The robot's design caters to a variety of applications, particularly emphasizing roles in caregiving and physical therapy. These application domains benefit from GR-1's stable and human-like movement which is essential in environments requiring precise and gentle interactions. == References == <references /> {{DEFAULTSORT:GR-1}} [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Fourier Intelligence Robotics]] 8ff8963b7958c44c80e40b0aa1541d0d03761be1 559 558 2024-04-28T00:10:45Z Stompy 14 Undo revision 558 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki GR-1 is a self-developed and mass-produced humanoid robot by [[Fourier Intelligence]]. It has a highly bionic torso and human-like motion control capabilities, with up to 54 Degrees of Freedom (DoFs) across its form. GR-1 can walk briskly, adroitly avoid obstacles, stably descend a slope, and withstand disruptions, making it an ideal physical agent of artificial general intelligence (AGI). {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | height = 165 cm | weight = 55 kg | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 | cost = USD 149,999 }} [[Category:Robots]] [[Category:Stompy, Expand!]] 46127ec7215edbbeaec4c6ce5e0eab82fccfe914 560 559 2024-04-28T00:11:40Z Stompy 14 Bot expanded article wikitext text/x-wiki The GR-1 is a highly sophisticated humanoid robot designed and mass-produced by [[Fourier Intelligence]]. With its highly bionic torso and human-like motion control capabilities, it initiates a new era in the field of robotics. The GR-1 features up to 54 Degrees of Freedom (DoFs) throughout its form, making it capable of performing complex actions such as brisk walking, adeptly avoiding obstacles, stably descending slopes, and withstanding disruptions<ref>https://fourierintelligence.com/gr1/</ref>. Built with a peak joint torque of 230N.m, GR-1 shows a considerable strength, enabling it to carry items nearly as heavy as its own weight<ref>https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref>. Its innovative design, combined with its advanced capabilities, puts it on the forefront of artificial general intelligence (AGI) physical embodiments. Fourier Intelligence plans to manufacture 100 of its GR-1 humanoid robots by the end of 2023<ref>https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref>. The mass production of this complex robot is a testament to the growing capabilities of robotics manufacturing and an indicator of how humanoid robots are becoming more widely accessible. == Specifications == The GR-1 measures 1.65m in height and weights 55kg<ref>https://www.therobotreport.com/fourier-intelligence-launches-production-version-of-gr-1-humanoid-robot/</ref>. The robot's joint module at the hip is specially designed for enhanced mobility and stability. These measurements, combined with its DoFs and AI capabilities, equip the GR-1 to perform tasks with human-like proficiency. == Affordability == The GR-1 humanoid robot is priced at USD 149,999, making it a significant investment. However, considering its advanced features and capabilities, it provides immense value for researchers, industries, and possibly consumers in the future. {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | cost = USD 149,999 | height = 165 cm | weight = 55 kg | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = 100 (planned by end of 2023) | dof = 54 | status = | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 }} [[Category:Robots]] == References == <references /> 694c4d310ff6049fd55cb9d38fda5485b9d7bdfd 561 560 2024-04-28T00:12:41Z Stompy 14 wikitext text/x-wiki {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | cost = USD 149,999 | height = 165 cm | weight = 55 kg | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = 100 (planned by end of 2023) | dof = 54 | status = | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 }} The GR-1 is a highly sophisticated humanoid robot designed and mass-produced by [[Fourier Intelligence]]. With its highly bionic torso and human-like motion control capabilities, it initiates a new era in the field of robotics. The GR-1 features up to 54 Degrees of Freedom (DoFs) throughout its form, making it capable of performing complex actions such as brisk walking, adeptly avoiding obstacles, stably descending slopes, and withstanding disruptions<ref>https://fourierintelligence.com/gr1/</ref>. Built with a peak joint torque of 230N.m, GR-1 shows a considerable strength, enabling it to carry items nearly as heavy as its own weight<ref>https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref>. Its innovative design, combined with its advanced capabilities, puts it on the forefront of artificial general intelligence (AGI) physical embodiments. Fourier Intelligence plans to manufacture 100 of its GR-1 humanoid robots by the end of 2023<ref>https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref>. The mass production of this complex robot is a testament to the growing capabilities of robotics manufacturing and an indicator of how humanoid robots are becoming more widely accessible. == Specifications == The GR-1 measures 1.65m in height and weights 55kg<ref>https://www.therobotreport.com/fourier-intelligence-launches-production-version-of-gr-1-humanoid-robot/</ref>. The robot's joint module at the hip is specially designed for enhanced mobility and stability. These measurements, combined with its DoFs and AI capabilities, equip the GR-1 to perform tasks with human-like proficiency. == Affordability == The GR-1 humanoid robot is priced at USD 149,999, making it a significant investment. However, considering its advanced features and capabilities, it provides immense value for researchers, industries, and possibly consumers in the future. == References == <references /> [[Category:Robots]] [[Category:Stompy, Expand!]] 64e758b5e6e2904de9cc810fbc4f9d0b76f7ea23 562 561 2024-04-28T00:14:18Z Stompy 14 Bot expanded article wikitext text/x-wiki {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | cost = USD 149,999 | height = 165 cm | weight = 55 kg | speed = 5 km/h<ref name="speed">https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref> | lift_force = Able to lift nearly its own weight<ref name="lift">https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref> | battery_life = Information not available | battery_capacity = Information not available | purchase_link = Information not available | number_made = 100 (planned by end of 2023) | dof = 54 | status = In mass-production<ref name="status">https://finance.yahoo.com/news/chinese-startup-debuts-world-first-130000324.html</ref> | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 }} The GR-1 is a highly sophisticated humanoid robot designed and mass-produced by [[Fourier Intelligence]]. With its highly bionic torso and human-like motion control capabilities, it initiates a new era in the field of robotics. The GR-1 features up to 54 Degrees of Freedom (DoFs) throughout its form, allowing it to perform complex actions such as brisk walking at around 5 km/h, adeptly avoiding obstacles, and stably descending slopes<ref name="speed"></ref>. The GR-1 robot has been designed to withstanding disruptions<ref>https://fourierintelligence.com/gr1/</ref>. Equipped with a peak joint torque of 230N.m, GR-1 demonstrates considerable strength, enabling it to carry items nearly as heavy as its own weight<ref name="lift"></ref>. Its innovative design, combined with its advanced capabilities, puts it on the forefront of artificial general intelligence (AGI) physical embodiments. Fourier Intelligence plans to manufacture 100 of its GR-1 humanoid robots by the end of 2023<ref>https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref>. The mass-production of this complex robot is a testament to the growing capabilities of robotics manufacturing, indicating the increasing accessibility of humanoid robots. == Specifications == The GR-1 measures 1.65m in height and weights 55kg<ref>https://www.therobotreport.com/fourier-intelligence-launches-production-version-of-gr-1-humanoid-robot/</ref>. Its joint module at the hip is specially designed for enhanced mobility and stability. These measurements, combined with its 54 DoFs and AI capabilities, equip the GR-1 to perform tasks with human-like proficiency. == Affordability == Priced at USD 149,999, the GR-1 humanoid robot is a significant investment. However, considering its advanced features and capabilities, it offers immense value for researchers, industries, and potentially consumers in the future. == References == <references /> [[Category:Robots]] a53229fd23f4c81b206abc193ed117415faac5b8 0 57 563 562 2024-04-28T00:14:48Z Stompy 14 wikitext text/x-wiki {{infobox robot | name = GR-1 | organization = [[Fourier Intelligence]] | cost = USD 149,999 | height = 165 cm | weight = 55 kg | speed = 5 km/h<ref name="speed">https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref> | lift_force = Able to lift nearly its own weight<ref name="lift">https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref> | battery_life = | battery_capacity = | purchase_link = | number_made = 100 (planned by end of 2023) | dof = 54 | status = In mass-production<ref name="status">https://finance.yahoo.com/news/chinese-startup-debuts-world-first-130000324.html</ref> | video_link = https://www.youtube.com/watch?v=SHPxcRBlXN0 }} The GR-1 is a highly sophisticated humanoid robot designed and mass-produced by [[Fourier Intelligence]]. With its highly bionic torso and human-like motion control capabilities, it initiates a new era in the field of robotics. The GR-1 features up to 54 Degrees of Freedom (DoFs) throughout its form, allowing it to perform complex actions such as brisk walking at around 5 km/h, adeptly avoiding obstacles, and stably descending slopes<ref name="speed"></ref>. The GR-1 robot has been designed to withstanding disruptions<ref>https://fourierintelligence.com/gr1/</ref>. Equipped with a peak joint torque of 230N.m, GR-1 demonstrates considerable strength, enabling it to carry items nearly as heavy as its own weight<ref name="lift"></ref>. Its innovative design, combined with its advanced capabilities, puts it on the forefront of artificial general intelligence (AGI) physical embodiments. Fourier Intelligence plans to manufacture 100 of its GR-1 humanoid robots by the end of 2023<ref>https://newatlas.com/robotics/fourier-gr1-humanoid-robot/</ref>. The mass-production of this complex robot is a testament to the growing capabilities of robotics manufacturing, indicating the increasing accessibility of humanoid robots. == Specifications == The GR-1 measures 1.65m in height and weights 55kg<ref>https://www.therobotreport.com/fourier-intelligence-launches-production-version-of-gr-1-humanoid-robot/</ref>. Its joint module at the hip is specially designed for enhanced mobility and stability. These measurements, combined with its 54 DoFs and AI capabilities, equip the GR-1 to perform tasks with human-like proficiency. == Affordability == Priced at USD 149,999, the GR-1 humanoid robot is a significant investment. However, considering its advanced features and capabilities, it offers immense value for researchers, industries, and potentially consumers in the future. == References == <references /> [[Category:Robots]] 725ccdbb378afcaedff5b45947674237ef0dd312 0 90 564 391 2024-04-28T00:16:58Z Stompy 14 wikitext text/x-wiki The [https://github.com/atopile/spin-servo-drive SPIN Servo] as an open-source actuator developed by [https://atopile.io/ atopile]. [[File:Spin.jpg|thumb]] {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque = 0.375nm | weight = 311.6g | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c }} [[Category:Actuators]] [[Category:Stompy, Expand!]] c882ffdbb4ccead331837b812dbb40ca1ce367fd 565 564 2024-04-28T00:17:28Z Stompy 14 Bot expanded article wikitext text/x-wiki == Overview == The SPIN Servo is an open-source hardware project, designed to make the use of fully-fledged Brushless DC (BLDC) servo motors easy and cost-effective<ref>[https://github.com/atopile/spin-servo-drive GitHub - atopile/spin-servo-drive: SPIN - Servos are awesome]</ref>. It is primarily engineered by atopile, which is known for its toolchains to describe electronic circuit boards with code<ref>[https://atopile.io/spin/ SPIN - atopile]</ref>. The intention behind this project is to introduce software development workflows like reuse, validation, and automation into the world of electronics. The SPIN Servo is manufactured by the Holry Motor company. It weighs 311.6g and its dimensions are 42mm x 42mm x 60mm. The cost of the Bill of Materials (BOM) is USD 30. The nominal torque of the spin servo is 0.125nm, while it can peak at 0.375nm. The interface employed by this servo is CAN bus and i2c. It operates at a voltage of 12V-24V and its gear ratio is directly driven, although there are bolt-on options available. All designs and schematics related to the SPIN Servo can be found at their official GitHub repository<ref>[https://github.com/atopile/spin-servo-drive/tree/main/mech SPIN Servo CAD - GitHub]</ref>. Interested individuals can purchase the SPIN Servo from the official atopile shop<ref>[https://shop.atopile.io/ Atopile Shop]</ref>. {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque = 0.375nm | weight = 311.6g | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c }} [[Category:Actuators]] == References == <references /> 605bbc904b4b9119bef1a57493aed73370b9ed5c 566 565 2024-04-28T00:18:33Z Stompy 14 wikitext text/x-wiki {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque = 0.375nm | weight = 311.6g | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c }} == Overview == The SPIN Servo is an open-source hardware project, designed to make the use of fully-fledged Brushless DC (BLDC) servo motors easy and cost-effective<ref>[https://github.com/atopile/spin-servo-drive GitHub - atopile/spin-servo-drive: SPIN - Servos are awesome]</ref>. It is primarily engineered by atopile, which is known for its toolchains to describe electronic circuit boards with code<ref>[https://atopile.io/spin/ SPIN - atopile]</ref>. The intention behind this project is to introduce software development workflows like reuse, validation, and automation into the world of electronics. The SPIN Servo is manufactured by the Holry Motor company. It weighs 311.6g and its dimensions are 42mm x 42mm x 60mm. The cost of the Bill of Materials (BOM) is USD 30. The nominal torque of the spin servo is 0.125nm, while it can peak at 0.375nm. The interface employed by this servo is CAN bus and i2c. It operates at a voltage of 12V-24V and its gear ratio is directly driven, although there are bolt-on options available. All designs and schematics related to the SPIN Servo can be found at their official GitHub repository<ref>[https://github.com/atopile/spin-servo-drive/tree/main/mech SPIN Servo CAD - GitHub]</ref>. Interested individuals can purchase the SPIN Servo from the official atopile shop<ref>[https://shop.atopile.io/ Atopile Shop]</ref>. == References == <references /> [[Category:Actuators]] a841dc657dbc4beb8f2634a45834612233b7fe2a 0 76 567 320 2024-04-28T00:19:53Z Stompy 14 wikitext text/x-wiki Deep Robotics is building a humanoid robot called [[Wukong-IV]]. {{infobox company | name = Deep Robotics | country = China | website_link = https://www.deeprobotics.cn/en | robots = [[Wukong-IV]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] d9d1dc40d83ea51b214f3876661a1b080528efe7 568 567 2024-04-28T00:20:25Z Stompy 14 Bot expanded article wikitext text/x-wiki Deep Robotics is a Chinese company known for its development of humanoid robots, including the [[Wukong-IV]] humanoid robot. As a leader in global quadruped robotics, Deep Robotics aims to serve the education, research, and entertainment markets with its advanced robotic technology<ref>DEEP Robotics showcases new X30 quadruped robot at GITEX GLOBAL 2023. (https://www.deeprobotics.cn/en/)</ref><ref>DEEP Robotics Launches Lite3 Quadruped Robot With Front Flipping ... (https://www.robotics247.com/article/deep_robotics_launches_lite3_quadruped_robot_front_flipping_mobility)</ref>. The company is not just limited to humanoid robots, it's also involved in the design and application of quadrupedal robots, including state-of-the-art names like Lite3 and X30. These robots can maneuver through various obstacles, navigate stairs, and function in extreme weather and lighting conditions, with applications in industries such as power, factory, pipeline, and research fields<ref>X30 is a bionic quadruped robot that can traverse obstacles. (https://www.therobotreport.com/deep-robotics-launches-x30-quadruped-robot-to-meet-varied-industry-needs/)</ref>. == Wukong-IV == [[Wukong-IV|Wukong-IV]], the humanoid robot developed by Deep Robotics, stands at 1.4 meters, weighs around 45 kg, and is actuated by electric motor joints. It features 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm. As a bionic humanoid robot, Wukong-IV's joint configuration and is capable of maintaining balance and stepping in place<ref>Hybrid Bipedal Locomotion Based on Reinforcement Learning and Heuristics. (https://pdfs.semanticscholar.org/f4d3/80f8e0fe39906f21f5270ffd2bf7bae74039.pdf)</ref><ref>Toward Understanding Key Estimation in Learning Robust Humanoid Locomotion. (https://arxiv.org/html/2403.05868v1)</ref>. {{infobox company | name = Deep Robotics | country = China | website_link = https://www.deeprobotics.cn/en | robots = [[Wukong-IV]] }} == References == <references /> [[Category:Companies]] a80aa36ff4237b276a0c1f9c08ed87b8ca2fbd31 569 568 2024-04-28T00:20:46Z Stompy 14 wikitext text/x-wiki {{infobox company | name = Deep Robotics | country = China | website_link = https://www.deeprobotics.cn/en | robots = [[Wukong-IV]] }} Deep Robotics is a Chinese company known for its development of humanoid robots, including the [[Wukong-IV]] humanoid robot. As a leader in global quadruped robotics, Deep Robotics aims to serve the education, research, and entertainment markets with its advanced robotic technology<ref>DEEP Robotics showcases new X30 quadruped robot at GITEX GLOBAL 2023. (https://www.deeprobotics.cn/en/)</ref><ref>DEEP Robotics Launches Lite3 Quadruped Robot With Front Flipping ... (https://www.robotics247.com/article/deep_robotics_launches_lite3_quadruped_robot_front_flipping_mobility)</ref>. The company is not just limited to humanoid robots, it's also involved in the design and application of quadrupedal robots, including state-of-the-art names like Lite3 and X30. These robots can maneuver through various obstacles, navigate stairs, and function in extreme weather and lighting conditions, with applications in industries such as power, factory, pipeline, and research fields<ref>X30 is a bionic quadruped robot that can traverse obstacles. (https://www.therobotreport.com/deep-robotics-launches-x30-quadruped-robot-to-meet-varied-industry-needs/)</ref>. == Wukong-IV == [[Wukong-IV|Wukong-IV]], the humanoid robot developed by Deep Robotics, stands at 1.4 meters, weighs around 45 kg, and is actuated by electric motor joints. It features 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm. As a bionic humanoid robot, Wukong-IV's joint configuration and is capable of maintaining balance and stepping in place<ref>Hybrid Bipedal Locomotion Based on Reinforcement Learning and Heuristics. (https://pdfs.semanticscholar.org/f4d3/80f8e0fe39906f21f5270ffd2bf7bae74039.pdf)</ref><ref>Toward Understanding Key Estimation in Learning Robust Humanoid Locomotion. (https://arxiv.org/html/2403.05868v1)</ref>. == References == <references /> [[Category:Companies]] c1d18362e76ccefb4c80f6e44927b57096feeacb 570 569 2024-04-28T00:20:55Z Stompy 14 wikitext text/x-wiki Deep Robotics is a Chinese company known for its development of humanoid robots, including the [[Wukong-IV]] humanoid robot. As a leader in global quadruped robotics, Deep Robotics aims to serve the education, research, and entertainment markets with its advanced robotic technology<ref>DEEP Robotics showcases new X30 quadruped robot at GITEX GLOBAL 2023. (https://www.deeprobotics.cn/en/)</ref><ref>DEEP Robotics Launches Lite3 Quadruped Robot With Front Flipping ... (https://www.robotics247.com/article/deep_robotics_launches_lite3_quadruped_robot_front_flipping_mobility)</ref>. {{infobox company | name = Deep Robotics | country = China | website_link = https://www.deeprobotics.cn/en | robots = [[Wukong-IV]] }} The company is not just limited to humanoid robots, it's also involved in the design and application of quadrupedal robots, including state-of-the-art names like Lite3 and X30. These robots can maneuver through various obstacles, navigate stairs, and function in extreme weather and lighting conditions, with applications in industries such as power, factory, pipeline, and research fields<ref>X30 is a bionic quadruped robot that can traverse obstacles. (https://www.therobotreport.com/deep-robotics-launches-x30-quadruped-robot-to-meet-varied-industry-needs/)</ref>. == Wukong-IV == [[Wukong-IV|Wukong-IV]], the humanoid robot developed by Deep Robotics, stands at 1.4 meters, weighs around 45 kg, and is actuated by electric motor joints. It features 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm. As a bionic humanoid robot, Wukong-IV's joint configuration and is capable of maintaining balance and stepping in place<ref>Hybrid Bipedal Locomotion Based on Reinforcement Learning and Heuristics. (https://pdfs.semanticscholar.org/f4d3/80f8e0fe39906f21f5270ffd2bf7bae74039.pdf)</ref><ref>Toward Understanding Key Estimation in Learning Robust Humanoid Locomotion. (https://arxiv.org/html/2403.05868v1)</ref>. == References == <references /> [[Category:Companies]] e2db3afc66107440a7c5d465fc0adb6c61b503f7 0 75 571 457 2024-04-28T00:21:21Z Stompy 14 wikitext text/x-wiki Wukong-IV is an adult-size humanoid robot designed and built by the research team at [[Deep Robotics]]. It is actuated by electric motor joints. The robot has 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm. {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_payload | two_hand_payload | cost = | video_link = https://www.youtube.com/watch?v=fbk4fYc6U14 }} [[Category:Robots]] [[Category:Stompy, Expand!]] 941fdbffe9abda8eecddd4420419833db77ddc9f 572 571 2024-04-28T00:21:48Z Stompy 14 Bot expanded article wikitext text/x-wiki I was unable to find information online about the cost of the Wukong-IV robot and its payload capabilities. However, we do have details about its height, weight, degrees of freedom, and that it is actuated by 21 electric motor joints. Here is the updated information: {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_payload = | two_hand_payload = | cost = | video_link = https://www.youtube.com/watch?v=fbk4fYc6U14 | dof = 6 (each leg), 4 (each arm) | number_made = | lift_force = | battery_life = | battery_capacity = | purchase_link = | status = | speed = }} The Wukong-IV is an adult-size humanoid robot designed and built by the research team at [[Deep Robotics]]. It stands 1.4 meters tall and weighs 45 kg<ref>https://pdfs.semanticscholar.org/f4d3/80f8e0fe39906f21f5270ffd2bf7bae74039.pdf</ref>. This bionic humanoid robot is actuated by 21 electric motor joints. It has 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm<ref>https://www.mdpi.com/2072-666X/13/10/1688</ref>. == References == <references /> [[Category:Robots]] 9ef0be8972f48ea64b27cf76b4333e87e07140bf 573 572 2024-04-28T00:22:15Z Stompy 14 wikitext text/x-wiki {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_payload = | two_hand_payload = | cost = | video_link = https://www.youtube.com/watch?v=fbk4fYc6U14 | dof = 6 (each leg), 4 (each arm) | number_made = | lift_force = | battery_life = | battery_capacity = | purchase_link = | status = | speed = }} The Wukong-IV is an adult-size humanoid robot designed and built by the research team at [[Deep Robotics]]. It stands 1.4 meters tall and weighs 45 kg<ref>https://pdfs.semanticscholar.org/f4d3/80f8e0fe39906f21f5270ffd2bf7bae74039.pdf</ref>. This bionic humanoid robot is actuated by 21 electric motor joints. It has 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm<ref>https://www.mdpi.com/2072-666X/13/10/1688</ref>. == References == <references /> [[Category:Robots]] 51aa24b00f8a0adcae33c3d53c2ac6c51ed59448 574 573 2024-04-28T00:22:30Z Stompy 14 wikitext text/x-wiki {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_payload = | two_hand_payload = | cost = | video_link = https://www.youtube.com/watch?v=fbk4fYc6U14 | dof = 21 | number_made = | lift_force = | battery_life = | battery_capacity = | purchase_link = | status = | speed = }} The Wukong-IV is an adult-size humanoid robot designed and built by the research team at [[Deep Robotics]]. It stands 1.4 meters tall and weighs 45 kg<ref>https://pdfs.semanticscholar.org/f4d3/80f8e0fe39906f21f5270ffd2bf7bae74039.pdf</ref>. This bionic humanoid robot is actuated by 21 electric motor joints. It has 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm<ref>https://www.mdpi.com/2072-666X/13/10/1688</ref>. == References == <references /> [[Category:Robots]] 39c28ca5d4674bde7be419cffd55f458c2e1fa30 0 106 575 412 2024-04-28T00:23:51Z Stompy 14 wikitext text/x-wiki MenteeBot is building a humanoid robot called [[MenteeBot (Robot)]]. {{infobox company | name = MenteeBot | country = Israel | website_link = https://www.menteebot.com/ | robots = [[MenteeBot (Robot)]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 1f00ea3dd9423249b35e36f262714c7d6211e9c8 576 575 2024-04-28T00:24:15Z Stompy 14 Bot expanded article wikitext text/x-wiki The information provided about the company MenteeBot is very concise and leaves several aspects uncovered. In order to expand this article, I will use the Tavily search API to find more information that we can include. uses(functions.search_web, { "query": "MenteeBot company" }).then(response => { Update article: MenteeBot is an Israel-based technology company that focuses on the design and construction of humanoid robots. Its primary creation is a humanoid robot also named [[MenteeBot (Robot)]]. The company aims to advance the development of robotics and human-robot interactions to foster a future where robots can efficiently assist humans in various tasks. The company was founded in (Year Found) with a mission to innovate and surpass the existing boundaries in robotics. It has since made considerable advancements towards the creation of efficient humanoid robots. {{infobox company | name = MenteeBot | country = Israel | website_link = https://www.menteebot.com/ | robots = [[MenteeBot (Robot)]] }} == Overview == MenteeBot was built around the concept of creating humanoid robots that can help people in their daily lives. These robots are designed to learn from human behavior and adapt to their needs. The robot MenteeBot is the company's main product and the epitome of its technological expertise and vision. == Notable Projects == The company's most noteworthy project is the [[MenteeBot (Robot)]], a humanoid robot model that has received recognition for its design and functionalities. == See also == * [[Humanoid robots]] * [[Robotics in Israel]] [[Category:Companies]] [[Category:Robot Manufacturers]] [[Category:Technology Companies in Israel]] == References == <references /> }).catch(error => { // handle error console.error(error); }); 03b7d70156d08cf42ca9e188bb5e6a06e2b3e412 577 576 2024-04-28T00:24:54Z Stompy 14 Undo revision 576 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki MenteeBot is building a humanoid robot called [[MenteeBot (Robot)]]. {{infobox company | name = MenteeBot | country = Israel | website_link = https://www.menteebot.com/ | robots = [[MenteeBot (Robot)]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 1f00ea3dd9423249b35e36f262714c7d6211e9c8 581 577 2024-04-28T06:44:51Z 108.211.178.220 0 wikitext text/x-wiki MenteeBot is building a humanoid robot called [[MenteeBot (Robot)]]. {{infobox company | name = MenteeBot | country = Israel | website_link = https://www.menteebot.com/ | robots = [[MenteeBot (Robot)]] }} [[Category:Companies]] d6961a3aab5a6bf59fcdcd5d0ac420d885dfcc71 2 149 578 2024-04-28T04:43:37Z Allen12 15 Created page with "{{infobox person | name = Allen Wu | organization = [[K-Scale Labs]] | title = Employee }} [[Category: K-Scale Employees]]" wikitext text/x-wiki {{infobox person | name = Allen Wu | organization = [[K-Scale Labs]] | title = Employee }} [[Category: K-Scale Employees]] 41b3ccf41dcb4839bd1fdcb58fdf2f9c36da0cf6 0 3 579 550 2024-04-28T06:44:15Z 108.211.178.220 0 wikitext text/x-wiki '''Unitree H1''' is a full-size universal humanoid robot developed by the [[Unitree]], a company known for its innovative robotic designs. The H1 is celebrated for its superior power performance capabilities and advanced powertrain technologies. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | height = 180 cm | weight = 47 kg | speed = >3.3 m/s | lift_force = | battery_life = | battery_capacity = 864 Wh | purchase_link = https://shop.unitree.com/products/unitree-h1 | number_made = | dof = | status = }} == Specifications == The H1 robot stands approximately 180 cm tall and weighs around 47 kg, offering high mobility and physical capabilities. Some of the standout specifications of the H1 include: * Maximum speed: Exceeds 3.3 meters per second, a benchmark in robot mobility. * Weight: Approximately 47 kg. * Maximum joint torque: 360 N.m. * Battery capacity: 864 Wh, which is quickly replaceable, enhancing the robot's operational endurance. == Features == The H1 incorporates advanced technologies to achieve its high functionality: * Highly efficient powertrain for superior speed, power, and maneuverability. * Equipped with high-torque joint motors developed by [[Unitree]] itself. * 360° depth sensing capabilities combined with LIDAR and depth cameras for robust environmental perception. == Uses and Applications == While detailed use cases of Unitree H1 are not extensively documented, the robot's build and capabilities suggest its suitability in complicated tasks requiring human-like dexterity and strength which might include industrial applications, complex terrain navigation, and interactive tasks. f3c854cc88bf2e6afc40be54cdba94be07fdbc6f 612 579 2024-04-28T18:46:50Z 185.169.0.227 0 wikitext text/x-wiki '''Unitree H1''' is a full-size universal humanoid robot developed by the [[Unitree]], a company known for its innovative robotic designs. The H1 is celebrated for its superior power performance capabilities and advanced powertrain technologies. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | height = 180 cm | weight = 47 kg | speed = >3.3 m/s | lift_force = | battery_life = | battery_capacity = 864 Wh | purchase_link = https://shop.unitree.com/products/unitree-h1 | number_made = | dof = | status = }} == Specifications == The H1 robot stands approximately 180 cm tall and weighs around 47 kg, offering high mobility and physical capabilities. Some of the standout specifications of the H1 include: * Maximum speed: Exceeds 3.3 meters per second, a benchmark in robot mobility. * Weight: Approximately 47 kg. * Maximum joint torque: 360 N.m. * Battery capacity: 864 Wh, which is quickly replaceable, enhancing the robot's operational endurance. == Features == The H1 incorporates advanced technologies to achieve its high functionality: * Highly efficient powertrain for superior speed, power, and maneuverability. * Equipped with high-torque joint motors developed by [[Unitree]] itself. * 360° depth sensing capabilities combined with LIDAR and depth cameras for robust environmental perception. == Uses and Applications == While detailed use cases of Unitree H1 are not extensively documented, the robot's build and capabilities suggest its suitability in complicated tasks requiring human-like dexterity and strength which might include industrial applications, complex terrain navigation, and interactive tasks. [[Category:Robots]] b9a87389aef0a24ad6c4967f4066a8a4dc42d1af 0 126 580 439 2024-04-28T06:44:39Z 108.211.178.220 0 wikitext text/x-wiki CyberOne is a humanoid robot from [[Xiaomi]]. {{infobox robot | name = CyberOne | organization = [[Xiaomi]] | height = 177 cm | weight = 52 kg | single_arm_payload = 1.5 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=yBmatGQ0giY | cost = }} [[Category:Robots]] [[Category:Stompy, Expand!]] 84807e69e223314d088276478494f64b777e220f 591 580 2024-04-28T07:04:44Z Stompy 14 Bot expanded article wikitext text/x-wiki CyberOne is a humanoid robot developed by the Chinese consumer electronics giant, Xiaomi. Unveiled in 2022 at a company event in Beijing by the founder, chairman, and CEO, Lei Jun, it is the newest member of Xiaomi's Cyber series, joining previously launched quadruped robots like CyberDog and CyberDog 2<ref>https://robotsguide.com/robots/cyberone</ref>. == Specifications == This bipedal humanoid robot has a height of 177 cm and weight of 52 kg, with an arm span of 168 cm<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. One of its distinct features is its ability to perceive 3D, recognize individuals, and respond to human emotions<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. Furthermore, it boasts a top speed of 3.6 km/ph<ref>https://www.theverge.com/2022/8/16/23307808/xiaomi-cyberone-humanoid-robot-tesla-optimus-bot-specs-comparison</ref>. == Pricing == The cost of CyberOne, if ever produced and made available for purchase, is estimated to be around 600,000 to 700,000 yuan<ref>https://robbreport.com/gear/electronics/xiaomi-humanoid-robot-cyberone-1234738597/</ref>. {{infobox robot | name = CyberOne | organization = [[Xiaomi]] | height = 177 cm | weight = 52 kg | single_arm_payload = 1.5 | runtime = | walk_speed = 3.6 km/h | video_link = https://www.youtube.com/watch?v=yBmatGQ0giY | cost = 600,000 - 700,000 yuan (est.) }} [[Category:Robots]] [[Category:Humanoid Robots]] == References == <references /> cee5154088296004b3baa81505eaa75b5413f18a 599 591 2024-04-28T07:44:43Z 192.145.118.30 0 wikitext text/x-wiki CyberOne is a humanoid robot developed by the Chinese consumer electronics giant, Xiaomi. Unveiled in 2022 at a company event in Beijing by the founder, chairman, and CEO, Lei Jun, it is the newest member of Xiaomi's Cyber series, joining previously launched quadruped robots like CyberDog and CyberDog 2<ref>https://robotsguide.com/robots/cyberone</ref>. {{infobox robot | name = CyberOne | organization = [[Xiaomi]] | height = 177 cm | weight = 52 kg | single_arm_payload = 1.5 | runtime = | walk_speed = 3.6 km/h | video_link = https://www.youtube.com/watch?v=yBmatGQ0giY | cost = 600,000 - 700,000 yuan (est.) }} == Specifications == This bipedal humanoid robot has a height of 177 cm and weight of 52 kg, with an arm span of 168 cm<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. One of its distinct features is its ability to perceive 3D, recognize individuals, and respond to human emotions<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. Furthermore, it boasts a top speed of 3.6 km/ph<ref>https://www.theverge.com/2022/8/16/23307808/xiaomi-cyberone-humanoid-robot-tesla-optimus-bot-specs-comparison</ref>. == Pricing == The cost of CyberOne, if ever produced and made available for purchase, is estimated to be around 600,000 to 700,000 yuan<ref>https://robbreport.com/gear/electronics/xiaomi-humanoid-robot-cyberone-1234738597/</ref>. [[Category:Robots]] [[Category:Humanoid Robots]] == References == <references /> 965e64b1869ec3959dd35ae96059ca0801a13f0a 0 125 582 468 2024-04-28T06:45:09Z 108.211.178.220 0 wikitext text/x-wiki The CL-1 is a humanoid robot developed by [[LimX Dynamics]]. It’s one of the few robots globally that can dynamically climb stairs based on real-time terrain perception. This advanced capability is attributed to [[LimX Dynamics]]’ motion control and AI algorithms, as well as its proprietary high-performing actuators and hardware system. {{infobox robot | name = CL-1 | organization = [[LimX Dynamics]] | height = | weight = | single_arm_payload = | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=sihIDeJ4Hmk | cost = }} [[Category:Robots]] [[Category:Stompy, Expand!]] 482a95e0b400e977c1c8b3e2ebd9a278eac799ee 590 582 2024-04-28T07:04:15Z Stompy 14 Bot expanded article wikitext text/x-wiki Apologies, but the specific specifications such as height, weight, speed, runtime, and cost of the CL-1 robot developed by LimX Dynamics are not readily available. The search results have not yielded this specific information at this time. More specific information may be obtainable through direct inquiry to the manufacturer, LimX Dynamics. Please note the CL-1 is one of the few robots globally that can dynamically climb stairs based on real-time terrain perception, an ability attributed to LimX Dynamics' advanced motion control and AI algorithms, proprietary high-performing actuators and a complex hardware system<ref>https://www.limxdynamics.com/en/humanoid-robot</ref><ref>https://medium.com/@limxdynamics/limx-dynamics-unveils-dynamic-testing-of-humanoid-robot-achieving-real-time-perceptive-stair-51d37b0cc6a5</ref>. == Reference == <references /> dcbb71cccc47d9a6c50f1a915308de69e5c11f81 598 590 2024-04-28T07:44:10Z 192.145.118.30 0 Undo revision 590 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki The CL-1 is a humanoid robot developed by [[LimX Dynamics]]. It’s one of the few robots globally that can dynamically climb stairs based on real-time terrain perception. This advanced capability is attributed to [[LimX Dynamics]]’ motion control and AI algorithms, as well as its proprietary high-performing actuators and hardware system. {{infobox robot | name = CL-1 | organization = [[LimX Dynamics]] | height = | weight = | single_arm_payload = | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=sihIDeJ4Hmk | cost = }} [[Category:Robots]] [[Category:Stompy, Expand!]] 482a95e0b400e977c1c8b3e2ebd9a278eac799ee 605 598 2024-04-28T08:04:09Z Stompy 14 Bot expanded article wikitext text/x-wiki The CL-1 is a humanoid robot developed by [[LimX Dynamics]]. It’s one of the few robots globally that can dynamically climb stairs based on real-time terrain perception. This advanced capability is attributed to [[LimX Dynamics]]’ motion control and AI algorithms, as well as its proprietary high-performing actuators and hardware system. Although precise specifications on the robot's height, weight, and cost are not publicly available at this time, it has been highlighted for its capabilities in navigating complex environments.<ref>https://www.therobotreport.com/limx-dynamics-shows-off-its-cl-1-humanoids-stair-climbing-abilities/</ref><ref>https://www.rockingrobots.com/limx-dynamics-shows-dynamic-testing-of-humanoid-robot-cl-1/</ref> {{infobox robot | name = CL-1 | organization = [[LimX Dynamics]] | height = Data not available | weight = Data not available | single_arm_payload = Data not available | runtime = Data not available | walk_speed = Data not available | video_link = https://www.youtube.com/watch?v=sihIDeJ4Hmk | cost = Data not available }} == Capabilities == The CL-1 robot is noted for its dynamic stair-climbing capabilities based on real-time terrain perception, a marked achievement in the realm of robotic development. This enables the robot to navigate complex environments, maneuvering indoors and outdoors, and effectively completing complex tasks such as climbing stairs and traversing slopes.<ref>https://www.therobotreport.com/limx-dynamics-shows-off-its-cl-1-humanoids-stair-climbing-abilities/</ref> == References == <references /> [[Category:Robots]] cff9b3c6c17e0a1dcddfda3baac0d94fc7ef65ee 0 138 583 476 2024-04-28T06:45:18Z 108.211.178.220 0 wikitext text/x-wiki Toyota Research Institute is building humanoid robot called [[T-HR3]]. {{infobox company | name = Toyota Research Institute | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] ea48212a16192cb22b976409e8f758ef3fe48e12 596 583 2024-04-28T07:07:42Z Stompy 14 Bot expanded article wikitext text/x-wiki == Toyota Research Institute == Toyota Research Institute (TRI) is a subsidiary of the multinational automotive manufacturer, Toyota, and is based in Japan. The institute is exploring various facets of robotics and is responsible for building Toyota's third generation humanoid robot, the [[T-HR3]]<ref>[https://spectrum.ieee.org/toyota-gets-back-into-humanoid-robots-with-new-thr3 Toyota Gets Back Into Humanoid Robots With New T-HR3]</ref>. == The T-HR3 Robot == The T-HR3 is a 1.5-meter humanoid robot designed to interact safely and naturally with its surroundings, and with the person controlling it<ref>[https://mag.toyota.co.uk/t-hr3-toyota-remote-control-robot/ T-HR3 - Toyota's new remote-controlled robot]</ref>. This robot weighs 75 kilograms and features 32 degrees of torque-controlled motion<ref>[https://spectrum.ieee.org/toyota-gets-back-into-humanoid-robots-with-new-thr3 Toyota Gets Back Into Humanoid Robots With New T-HR3]</ref>. It possesses some core capabilities: flexible joint control, whole-body coordination and balance control, and real remote maneuvering – thanks to its Torque Servo Module. These features enable the robot to control the force of contact it makes with any individual or object in its environment, to retain balance even when it collides with objects, and to mirror user movements with seamless and intuitive control<ref>[https://mag.toyota.co.uk/t-hr3-toyota-remote-control-robot/ T-HR3 - Toyota's new remote-controlled robot]</ref>. The development of the T-HR3 is championed by a research and development group led by Tomohisa Moridaira, based at Toyota's Tokyo Head Office.<ref>[https://global.toyota/en/newsroom/corporate/30609642.html Why is Toyota Developing Humanoid Robots?]</ref> == References == <references /> {{infobox company | name = Toyota Research Institute | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} [[Category:Companies]] [[Category:Humanoid Robots]] cd65073b53a1bb212f5c5ca29ca19c4418189ce7 603 596 2024-04-28T07:46:46Z 192.145.118.30 0 wikitext text/x-wiki Toyota Research Institute (TRI) is a subsidiary of the multinational automotive manufacturer, Toyota, and is based in Japan. The institute is exploring various facets of robotics and is responsible for building Toyota's third generation humanoid robot, the [[T-HR3]]<ref>[https://spectrum.ieee.org/toyota-gets-back-into-humanoid-robots-with-new-thr3 Toyota Gets Back Into Humanoid Robots With New T-HR3]</ref>. {{infobox company | name = Toyota Research Institute | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} == The T-HR3 Robot == The T-HR3 is a 1.5-meter humanoid robot designed to interact safely and naturally with its surroundings, and with the person controlling it<ref>[https://mag.toyota.co.uk/t-hr3-toyota-remote-control-robot/ T-HR3 - Toyota's new remote-controlled robot]</ref>. This robot weighs 75 kilograms and features 32 degrees of torque-controlled motion<ref>[https://spectrum.ieee.org/toyota-gets-back-into-humanoid-robots-with-new-thr3 Toyota Gets Back Into Humanoid Robots With New T-HR3]</ref>. It possesses some core capabilities: flexible joint control, whole-body coordination and balance control, and real remote maneuvering – thanks to its Torque Servo Module. These features enable the robot to control the force of contact it makes with any individual or object in its environment, to retain balance even when it collides with objects, and to mirror user movements with seamless and intuitive control<ref>[https://mag.toyota.co.uk/t-hr3-toyota-remote-control-robot/ T-HR3 - Toyota's new remote-controlled robot]</ref>. The development of the T-HR3 is championed by a research and development group led by Tomohisa Moridaira, based at Toyota's Tokyo Head Office.<ref>[https://global.toyota/en/newsroom/corporate/30609642.html Why is Toyota Developing Humanoid Robots?]</ref> == References == <references /> [[Category:Companies]] [[Category:Humanoid Robots]] b8a430ff64fb916e038bb8e5ae58a442e2cd875b 0 137 584 471 2024-04-28T06:45:31Z 108.211.178.220 0 wikitext text/x-wiki Punyo is a soft robot developed by the [[Toyota Research Institute]] (TRI) to revolutionize whole-body manipulation research. Unlike traditional robots that primarily use hands for manipulation, Punyo employs its arms and chest. The robot is designed to help with everyday tasks, such as lifting heavy objects or closing a drawer. {{infobox robot | name = Punyo | organization = [[Toyota Research Institute]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=FY-MD4gteeE | cost = }} [[Category:Robots]] [[Category:Stompy, Expand!]] c30f8b691f91c43e54508188502f4dfd70262978 593 584 2024-04-28T07:05:42Z Stompy 14 Bot expanded article wikitext text/x-wiki I wasn't able to find the specific height, weight, or cost of the Punyo robot. However, I was able to find general information about the robot, which employed whole-body manipulation, using its arms and chest, to assist with everyday tasks<ref>https://newatlas.com/robotics/toyota-punyo-humanoid-soft-robot/</ref><ref>https://punyo.tech/</ref>. The robot platform is constructed with compliant materials and tactile mechanisms that increase its ability to safely handle objects, especially large and heavy ones<ref>https://medium.com/toyotaresearch/meet-punyo-tris-soft-robot-for-whole-body-manipulation-research-949c934ac3d8</ref><ref>https://spectrum.ieee.org/humanoid-robot-tri-punyo</ref>. Here is the updated information for the Punyo robot: {{infobox robot | name = Punyo | organization = [[Toyota Research Institute]] | height = Unknown | weight = Unknown | two_hand_payload = Unknown | video_link = https://www.youtube.com/watch?v=FY-MD4gteeE | cost = Unknown }} == Article == Punyo is a soft robot developed by the [[Toyota Research Institute]] (TRI) to revolutionize whole-body manipulation research. Unlike traditional robots that primarily use hands for manipulation, Punyo employs its arms and chest. The robot is designed to help with everyday tasks, such as lifting heavy objects or closing a drawer<ref>https://newatlas.com/robotics/toyota-punyo-humanoid-soft-robot/</ref><ref>https://punyo.tech/</ref>. == Description == Punyo's platform designs with compliant materials and employs tactile mechanisms that increase its ability to handle objects effectively and safely, especially large and heavy ones<ref>https://medium.com/toyotaresearch/meet-punyo-tris-soft-robot-for-whole-body-manipulation-research-949c934ac3d8</ref><ref>https://spectrum.ieee.org/humanoid-robot-tri-punyo</ref>. [[Category:Robots]] == References == <references /> b034b3a39ff1ffbbbdb3d66d695a8b480ee8b340 601 593 2024-04-28T07:45:56Z 192.145.118.30 0 wikitext text/x-wiki I wasn't able to find the specific height, weight, or cost of the Punyo robot. However, I was able to find general information about the robot, which employed whole-body manipulation, using its arms and chest, to assist with everyday tasks<ref>https://newatlas.com/robotics/toyota-punyo-humanoid-soft-robot/</ref><ref>https://punyo.tech/</ref>. The robot platform is constructed with compliant materials and tactile mechanisms that increase its ability to safely handle objects, especially large and heavy ones<ref>https://medium.com/toyotaresearch/meet-punyo-tris-soft-robot-for-whole-body-manipulation-research-949c934ac3d8</ref><ref>https://spectrum.ieee.org/humanoid-robot-tri-punyo</ref>. Here is the updated information for the Punyo robot: {{infobox robot | name = Punyo | organization = [[Toyota Research Institute]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=FY-MD4gteeE | cost = }} == Article == Punyo is a soft robot developed by the [[Toyota Research Institute]] (TRI) to revolutionize whole-body manipulation research. Unlike traditional robots that primarily use hands for manipulation, Punyo employs its arms and chest. The robot is designed to help with everyday tasks, such as lifting heavy objects or closing a drawer<ref>https://newatlas.com/robotics/toyota-punyo-humanoid-soft-robot/</ref><ref>https://punyo.tech/</ref>. == Description == Punyo's platform designs with compliant materials and employs tactile mechanisms that increase its ability to handle objects effectively and safely, especially large and heavy ones<ref>https://medium.com/toyotaresearch/meet-punyo-tris-soft-robot-for-whole-body-manipulation-research-949c934ac3d8</ref><ref>https://spectrum.ieee.org/humanoid-robot-tri-punyo</ref>. == References == <references /> [[Category:Robots]] f3102b97c1415e924b918a5cddde764a19350801 0 111 585 417 2024-04-28T06:45:43Z 108.211.178.220 0 wikitext text/x-wiki PX5 is a humanoid robot from [[Xpeng]]. {{infobox robot | name = PX5 | organization = [[Xpeng]] | height = | weight = | single_hand_payload | two_hand_payload = | video_link = https://www.youtube.com/watch?v=BNSZ8Fwcd20 | cost }} [[Category:Robots]] [[Category:Stompy, Expand!]] cd88c74d1ebb4fcadf61b17b471cbe9bb4565b80 594 585 2024-04-28T07:07:04Z Stompy 14 Bot expanded article wikitext text/x-wiki The PX5 is a humanoid robot developed by Xpeng, unveiled for the first time during Xpeng Motors' Tech Day in 2023.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> The robot stands approximately 1.5 meters in height and is able to navigate different terrain and handle objects with precision, demonstrating remarkable stability.<ref>https://technode.com/2023/10/25/xpeng-tech-day-2023-first-mpv-self-driving-timeline-flying-cars-and-humanoid-robots/</ref> Constructed with a silver-white color scheme, the PX5 is also resistant to shock.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> {{infobox robot | name = PX5 | organization = [[Xpeng]] | height = 1.5 meters | weight = | video_link = https://www.youtube.com/watch?v=BNSZ8Fwcd20 | cost = }} == Development == Xpeng Robotics, an ecosystem company of Xpeng, specializing in smart robots, revealed the PX5. The company, which was founded in 2016, innovates in areas like robot powertrain, locomotion control, robot autonomy, robot interaction, and artificial intelligence contributing to a shared mission of exploring future mobility solutions.<ref>https://www.pxing.com/en/about</ref> == Design and Capabilities == The PX5 bears a striking silver-white finish and exhibits resistance to shock. Its capability to navigate through different terrains and handle handheld objects, such as a pen, with exceptional stability has been highlighted in demonstrations.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> <ref>https://technode.com/2023/10/25/xpeng-tech-day-2023-first-mpv-self-driving-timeline-flying-cars-and-humanoid-robots/</ref> [[Category:Robots]] [[Category:Xpeng Robots]] == References == <references /> c6b8a0a6adb3aac63cce35a967293da314182ab0 0 1 586 419 2024-04-28T06:46:02Z 108.211.178.220 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |} 152351df7d24c1b25a6bc27068461d60c15ebcb4 608 586 2024-04-28T16:23:39Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |} 9f488d3d39b8bb0b0469f8cf9ca556cbb3e52158 0 109 587 415 2024-04-28T06:46:19Z 108.211.178.220 0 wikitext text/x-wiki Nadia is a humanoid robot from [[IHMC, Boardwalk Robotics]]. {{infobox robot | name = Nadia | organization = [[IHMC, Boardwalk Robotics]] | height = | weight = | single_hand_payload | two_hand_payload = | video_link = https://www.youtube.com/watch?v=uTmUfOc7r_s | cost }} [[Category:Robots]] [[Category:Stompy, Expand!]] 20098c581d4658fba1d2676af685464e23a15abe 592 587 2024-04-28T07:05:10Z Stompy 14 Bot expanded article wikitext text/x-wiki Nadia is a humanoid robot developed by the Institute for Human & Machine Cognition (IHMC) in collaboration with Boardwalk Robotics<ref>[http://robots.ihmc.us/nadia Nadia Humanoid — IHMC Robotics Lab]</ref>. This advanced humanoid robot, named Nadia, is designed with a high power-to-weight ratio and a large range of motion, characteristics which provide it with exceptional mobility<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. Though specific information regarding height, weight, and payload capacities have not been explicitly stated, Nadia reportedly has one of the highest ranges of motion across its 29 joints of any humanoid robot globally<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. {{infobox robot | name = Nadia | organization = [[IHMC, Boardwalk Robotics]] | height = Not available | weight = Not available | single_hand_payload = Not available | two_hand_payload = Not available | video_link = https://www.youtube.com/watch?v=uTmUfOc7r_s | cost = Not available }} == Design and Capabilities == The Nadia humanoid robot's design encompasses a high power-to-weight ratio, contributing to its significant mobility potential. It stands out due to its extensive range of motion, facilitated by its architecture of 29 joints<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. These design features enable Nadia to adapt to and function within urban environments, aligning with the project's goal of facilitating semi-autonomous behaviors. Built with the same intelligence that powers the IHMC's DRC-Atlas robot, Nadia boasts real-time perception, compliant locomotion, autonomous footstep placement, and dexterous VR-teleoperated manipulation<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. == Research and Development == The development of Nadia is a collaborative project by the IHMC Robotics Lab and Boardwalk Robotics. The research team aims to produce a next-generation humanoid, capable of executing more perilous tasks while retaining high mobility<ref>[https://www.bbc.com/news/world-us-canada-67722014 A VR-controlled robot that throws boxing punches - BBC]</ref>. This development project positions Nadia as one of the most mobile ground robots designed in-house at IHMC in nearly a decade<ref>[https://www.ihmc.us/news20221005/ Video shows progress of IHMC humanoid robot Nadia]</ref>. == References == <references /> [[Category:Robots]] 46cf2b83efd2741aecdad111bdebba3a37205110 600 592 2024-04-28T07:45:19Z 192.145.118.30 0 wikitext text/x-wiki Nadia is a humanoid robot developed by the Institute for Human & Machine Cognition (IHMC) in collaboration with Boardwalk Robotics<ref>[http://robots.ihmc.us/nadia Nadia Humanoid — IHMC Robotics Lab]</ref>. This advanced humanoid robot, named Nadia, is designed with a high power-to-weight ratio and a large range of motion, characteristics which provide it with exceptional mobility<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. Though specific information regarding height, weight, and payload capacities have not been explicitly stated, Nadia reportedly has one of the highest ranges of motion across its 29 joints of any humanoid robot globally<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. {{infobox robot | name = Nadia | organization = [[IHMC, Boardwalk Robotics]] | height = | weight = | single_hand_payload = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=uTmUfOc7r_s | cost = }} == Design and Capabilities == The Nadia humanoid robot's design encompasses a high power-to-weight ratio, contributing to its significant mobility potential. It stands out due to its extensive range of motion, facilitated by its architecture of 29 joints<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. These design features enable Nadia to adapt to and function within urban environments, aligning with the project's goal of facilitating semi-autonomous behaviors. Built with the same intelligence that powers the IHMC's DRC-Atlas robot, Nadia boasts real-time perception, compliant locomotion, autonomous footstep placement, and dexterous VR-teleoperated manipulation<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. == Research and Development == The development of Nadia is a collaborative project by the IHMC Robotics Lab and Boardwalk Robotics. The research team aims to produce a next-generation humanoid, capable of executing more perilous tasks while retaining high mobility<ref>[https://www.bbc.com/news/world-us-canada-67722014 A VR-controlled robot that throws boxing punches - BBC]</ref>. This development project positions Nadia as one of the most mobile ground robots designed in-house at IHMC in nearly a decade<ref>[https://www.ihmc.us/news20221005/ Video shows progress of IHMC humanoid robot Nadia]</ref>. == References == <references /> [[Category:Robots]] aecd83d6eb23b9871acad3520b0184e3d5a1a5c3 0 116 588 429 2024-04-28T06:46:43Z 108.211.178.220 0 wikitext text/x-wiki Westwood Robotics is building a humanoid robot called [[THEMIS]]. {{infobox company | name = Westwood Robotics | country = USA | website_link = https://www.westwoodrobotics.io/ | robots = [[THEMIS]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 1b5ce03e8e2e32f9c6139992abda2505e303d0bf 597 588 2024-04-28T07:08:07Z Stompy 14 Bot expanded article wikitext text/x-wiki Westwood Robotics, an American company, is developing a humanoid robot, THEMIS. Inspired by the company's aspiration to embed versatile humanoid robots into our daily life, THEMIS demonstrates Westwood's commitment to developing robots capable of executing complex tasks with human-like grace and aptitude. THEMIS embodies Westwood's vision of a future where humanoid robots interact safely with their environment while displaying dynamic movements comparable to human motions. Designed to perform general-purpose tasks, THEMIS is optimized for dynamic movements, including walking, running, and even parkour maneuvers. Founded in 2018, Westwood Robotics was established by a collection of enthusiastic roboticists hailing from the Robotics & Mechanisms Laboratory (RoMeLa). Driven by a mission to bring robots closer to people, the organization seeks to improve the interaction between humans and robots, ensuring safer interactions and more versatile applications. == THEMIS Development == Westwood Robotics' THEMIS is an advanced humanoid robot engineered to interact with its environment safely while executing an extraordinary range of motions and maneuvers. Utilizing state-of-the-art technology, it is designed to perform tasks with a degree of precision and finesse that matches or surpasses human capabilities. Its design allows for complex and dynamic maneuvers including walking, running, and intricate parkour movements, testaments to the firm's commitment to revolutionize the potential applications of humanoid robotics[[1]]. == History == Westwood Robotics was founded in 2018 by a passionate group of roboticists from the Robotics & Mechanisms Laboratory (RoMeLa). With their vast expertise in robotics, they are dedicated to developing humanoid robots that can interact safely with humans and their environment. Their mission to bring robots closer to people is succinctly encapsulated in their flagship product, THEMIS[[2]]. == See Also == * [[Science and technology in the United States]] * [[List of robot companies]] == References == <references> <ref name="MicroStrainCaseStudy">[https://www.microstrain.com/blog/case-study-pushing-the-boundaries-of-humanoid-robot-performance-with-westwood Case Study: Pushing the Boundaries of Humanoid Robot Performance with Westwood Robotics THEMIS]</ref> <ref name="LinkedInProfile">[https://www.linkedin.com/company/westwoodrobotics Westwood Robotics Corporation LinkedIn Profile]</ref> </references> [[Category:Companies]] [[Category:Robotics Companies]] [[Category:Humanoid Robots]] [[Category:Science and technology in the United States]] 5f034d6daa0b1d8c36aa935b3ff26a35f2d55741 604 597 2024-04-28T07:47:04Z 192.145.118.30 0 Undo revision 597 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki Westwood Robotics is building a humanoid robot called [[THEMIS]]. {{infobox company | name = Westwood Robotics | country = USA | website_link = https://www.westwoodrobotics.io/ | robots = [[THEMIS]] }} [[Category:Companies]] [[Category:Stompy, Expand!]] 1b5ce03e8e2e32f9c6139992abda2505e303d0bf 607 604 2024-04-28T08:04:48Z Stompy 14 Bot expanded article wikitext text/x-wiki == Westwood Robotics == Westwood Robotics is an innovative company based in the USA, dedicated to bringing robots closer to people. Founded in 2018 by a team of passionate roboticists from the Robotics & Mechanisms Laboratory (RoMeLa) at University of California, Los Angeles (UCLA), Westwood Robotics is renowned for its advanced robotic applications.<ref>https://www.linkedin.com/company/westwoodrobotics</ref> The company is known for the development of diverse robotics products, among which is the THEMIS, an exceptional humanoid robot. == THEMIS == THEMIS is a humanoid robot developed by Westwood Robotics. The details about the product are not clear from available online sources. == Other Products == Apart from THEMIS, Westwood Robotics has been reported to have developed and introduced other products which include a kid-size humanoid robot used as an open-platform for research and education, as well as the BEAR series actuators that are known for exceptional live compliance and unparalleled dynamic performance.<ref>https://www.westwoodrobotics.io/</ref> {{infobox company | name = Westwood Robotics | country = USA | website_link = https://www.westwoodrobotics.io/ | robots = [[THEMIS]] }} [[Category:Companies]] == References == <references /> c20b21639a988f54b21fd1e95fc641dfb9488439 0 115 589 428 2024-04-28T06:46:51Z 108.211.178.220 0 wikitext text/x-wiki THEMIS is a humanoid robot from [[Westwood Robotics]]. {{infobox robot | name = THEMIS | organization = [[Westwood Robotics]] | height = 142.2 cm | weight = 39 kg | video_link = https://www.youtube.com/watch?v=yt4mHwAl9cc | cost = }} [[Category:Robots]] [[Category:Stompy, Expand!]] 90a63767a105f7059c3e345706b28ba6b2e78797 595 589 2024-04-28T07:07:17Z Stompy 14 Bot expanded article wikitext text/x-wiki The cost information for the THEMIS robot from Westwood Robotics is not readily available in the public domain. This could be for a number of reasons such as the robot's use in specific and restricted research domains, or the fluctuation of cost parameters based on customizations or updates to the model. Consequently, the cost is not specified in the infobox for the robot. The revised infobox would be: {{infobox robot | name = THEMIS | organization = [[Westwood Robotics]] | height = 142.2 cm | weight = 39 kg | video_link = https://www.youtube.com/watch?v=yt4mHwAl9cc | cost = Information not available in the public domain }} [[Category:Robots]] 88bc12827aff018d841a68c91e25cc15d8705817 602 595 2024-04-28T07:46:21Z 192.145.118.30 0 Undo revision 595 by [[Special:Contributions/Stompy|Stompy]] ([[User talk:Stompy|talk]]) wikitext text/x-wiki THEMIS is a humanoid robot from [[Westwood Robotics]]. {{infobox robot | name = THEMIS | organization = [[Westwood Robotics]] | height = 142.2 cm | weight = 39 kg | video_link = https://www.youtube.com/watch?v=yt4mHwAl9cc | cost = }} [[Category:Robots]] [[Category:Stompy, Expand!]] 90a63767a105f7059c3e345706b28ba6b2e78797 606 602 2024-04-28T08:04:29Z Stompy 14 Bot expanded article wikitext text/x-wiki While the search results have shown information about a robot named THEMIS, it seems to be about the THeMIS unmanned ground vehicle (UGV) developed by the company Milrem Robotics, as seen on the Milrem Robotics website<ref>[Milrem Robotics Defence](https://milremrobotics.com/defence/)</ref>. The search results do not provide additional information about the THEMIS humanoid robot from Westwood Robotics. Based on the initial information provided, we can update the infobox as follows (note that the cost field remains unfilled): {{infobox robot | name = THEMIS | organization = [[Westwood Robotics]] | video_link = https://www.youtube.com/watch?v=yt4mHwAl9cc | height = 142.2 cm | weight = 39 kg | cost = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} Further details about the robot's costs, speed, lift force, number made, degrees of freedom, battery life/capacity, and status are needed to complete the infobox. Until we are able to elaborate more on the THEMIS humanoid robot, we cannot finalize this article. == References == <references /> [[Category:Robots]] 2b3fe98a0a7d9d60acc83f54e83c388b39feb11c 0 150 609 2024-04-28T16:27:08Z Modeless 7 Created page with "Sharing the name of the Chinese space station for some reason, this robot is claimed to be the "first" running electric humanoid. Unitree H1 already runs faster though. Here i..." wikitext text/x-wiki Sharing the name of the Chinese space station for some reason, this robot is claimed to be the "first" running electric humanoid. Unitree H1 already runs faster though. Here is an article on the announcement with video: https://www.maginative.com/article/meet-tiangong-chinas-full-size-electric-running-humanoid-robot/ 00b08c40be9106acae944d05a6af6e27f5291755 0 151 610 2024-04-28T18:46:11Z 185.169.0.227 0 Created page with "Astribot is a Chinese bot from a subsidiary of [[Stardust Intelligence]]. [[Category:Stompy, Expand!]]" wikitext text/x-wiki Astribot is a Chinese bot from a subsidiary of [[Stardust Intelligence]]. [[Category:Stompy, Expand!]] c21f39c97f25f51c0953aebeb1d6ed6ecaa3d3f1 611 610 2024-04-28T18:46:34Z 185.169.0.227 0 wikitext text/x-wiki Astribot is a Chinese bot from a subsidiary of [[Stardust Intelligence]]. [[Category:Robots]] [[Category:Stompy, Expand!]] e52d4aebea0ed7651abddcfbd5c2655bc52ab48d 0 1 613 608 2024-04-28T18:47:12Z 185.169.0.227 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |} 1d04e36953efdcbf23c1abb2a170bd8fdff9b84c 627 613 2024-04-29T01:52:02Z 194.195.93.80 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |} 7f2b254cfcc3b6dc1bf50eb66c7aedfb437f40a8 629 627 2024-04-29T04:33:49Z 108.211.178.220 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |} 70ff00969755912c7f1f1bc940f7231e3076f27a 632 629 2024-04-29T16:12:19Z 38.108.138.43 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots ! License ! Links |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] | [[Open Source]] | [[https://github.com/assadollahi/kayra]] |} 8d6bbaebefb0ac5c5eadef1ac9edcfa99d2e054b 634 632 2024-04-29T17:41:27Z 185.187.168.105 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 95bcdfc7b05ed863b21d0b94f5ddba47997bf06a 0 151 614 611 2024-04-28T19:04:06Z Stompy 14 Bot expanded article wikitext text/x-wiki == Overview == Astribot, also known as the S1 robot, is a product of the Chinese tech company AstriBot Corporation, a subsidiary of Stardust Intelligence. The robot's name is derived from the Latin proverb "Ad astra per aspera," which translates to "through the arduous journey, to reach the stardust."<ref>https://www.maginative.com/article/stardust-intelligences-impressive-s1-robot-demo-raises-eyebrows/</ref> The Astribot showcases significant developments in the field of humanoid robotics. It features impressive performance specifications such as a movement speed of up to 10 meters per second and a lift ability up to 10 kilograms, marking a notable stride in AI-controlled humanoid robots.<ref>https://elblog.pl/2024/04/28/astribot-corporations-s1-robot-promises-swift-and-skilled-assistance/</ref> Astribot's parent company, Stardust Intelligence, was founded by Lai Jie, who also contributed to Tencent Robotics Laboratory, Baidu, and Hong Kong Polytechnic. {{infobox robot | name = Astribot | organization = AstriBot Corporation | video_link = | cost = | height = | weight = | speed = 10 m/s | lift_force = 10 kg | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} == Capabilities == Astribot S1 is described as an 'all-purpose' home robot. While the full extent of its capabilities are still being unveiled, it is said to have the potential for tasks like folding clothes, cooking meals, and cleaning rooms.<ref>https://businessnewsforkids.substack.com/p/85-a-super-cool-home-robot</ref> == References == <references /> [[Category:Robots]] [[Category:Chinese Robots]] [[Category:Stardust Intelligence]] b856fac9e84a7a6403844483b813afd311f744d1 615 614 2024-04-28T23:58:19Z 194.195.93.78 0 wikitext text/x-wiki == Overview == Astribot, also known as the S1 robot, is a product of the Chinese tech company [[AstriBot Corporation]], a subsidiary of Stardust Intelligence. The robot's name is derived from the Latin proverb "Ad astra per aspera," which translates to "through the arduous journey, to reach the stardust."<ref>https://www.maginative.com/article/stardust-intelligences-impressive-s1-robot-demo-raises-eyebrows/</ref> The Astribot showcases significant developments in the field of humanoid robotics. It features impressive performance specifications such as a movement speed of up to 10 meters per second and a lift ability up to 10 kilograms, marking a notable stride in AI-controlled humanoid robots.<ref>https://elblog.pl/2024/04/28/astribot-corporations-s1-robot-promises-swift-and-skilled-assistance/</ref> Astribot's parent company, Stardust Intelligence, was founded by Lai Jie, who also contributed to Tencent Robotics Laboratory, Baidu, and Hong Kong Polytechnic. {{infobox robot | name = Astribot | organization = AstriBot Corporation | video_link = | cost = | height = | weight = | speed = 10 m/s | lift_force = 10 kg | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} == Capabilities == Astribot S1 is described as an 'all-purpose' home robot. While the full extent of its capabilities are still being unveiled, it is said to have the potential for tasks like folding clothes, cooking meals, and cleaning rooms.<ref>https://businessnewsforkids.substack.com/p/85-a-super-cool-home-robot</ref> == References == <references /> [[Category:Robots]] [[Category:Stompy, Expand!]] df461eb4bf551f5ec39fe06d07524fcf17248ac2 619 615 2024-04-29T00:04:41Z Stompy 14 Bot expanded article wikitext text/x-wiki The Astribot S1 boasts an array of features that sets it apart from its predecessors. In its design, it intended to operate without the need for human guidance amidst its tasks, further emphasizing its autonomous ability<ref>https://www.geeky-gadgets.com/astribot-s1-ai-humanoid-robot/</ref>. The robot is known to assist in a variety of housework, including preparing drinks and doing other household chores such as ironing and folding<ref>https://www.msn.com/en-us/news/technology/astribot-ai-powered-humanoid-torso-can-prepare-drinks-help-with-housework/ar-AA1nP0ah</ref>. Key specifications of the Astribot S1 include a top speed of 10 meters per second, which is faster than the average adult man, a payload capacity of 10 kg per arm and a total of 7 degrees of freedom per arm. These arms exhibit a range of motion rivaling that of a human limb<ref>https://digialps.com/meet-the-incredibly-fast-astribot-s1-the-humanoid-robot-that-learns-at-1x-speed-without-any-help/</ref>. {{infobox robot | name = Astribot | organization = AstriBot Corporation | video_link = | cost = | height = | weight = | speed = 10 m/s | lift_force = 10 kg | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = 7 (per arm) | status = }} == Capabilities == Astribot S1 is described as an 'all-purpose' home robot. While the full extent of its capabilities are still being unveiled, it is said to have the potential for tasks like folding clothes, cooking meals, and cleaning rooms<ref>https://businessnewsforkids.substack.com/p/85-a-super-cool-home-robot</ref>. In addition to these home tasks, the Astribot S1 is also adept at performing intricate tasks such as opening bottles and pouring drinks, which clearly illustrates the robot's well-designed dexterity<ref>https://www.msn.com/en-us/news/technology/astribot-ai-powered-humanoid-torso-can-prepare-drinks-help-with-housework/ar-AA1nP0ah</ref>. == References == <references /> [[Category:Robots]] 38c35a87800eb4174cec1590a041c5f1569a931d 0 152 616 2024-04-28T23:58:39Z 194.195.93.78 0 Created page with "Creators of the [[Astribot S1]]. [[Category:Stompy, Expand!]]" wikitext text/x-wiki Creators of the [[Astribot S1]]. [[Category:Stompy, Expand!]] 01587f1632781104c70ce791d728d9514c942c3f 617 616 2024-04-28T23:58:57Z 194.195.93.78 0 wikitext text/x-wiki Creators of the [[Astribot S1]]. [[Category:Companies]] [[Category:Stompy, Expand!]] 2f1a57e6682f5a60a7ea4e8ec9ece55d513af156 618 617 2024-04-29T00:04:12Z Stompy 14 Bot expanded article wikitext text/x-wiki The Astribot S1 is a product of the Chinese tech company Astribot Corporation, a subsidiary of Stardust Intelligence. This firm has made notable strides in the field of AI-controlled humanoid robots, developing the Astribot S1 as a highly efficient helper capable of lifting objects weighing up to 10 kilograms and moving at a speed of 10 meters per second.<ref>https://elblog.pl/2024/04/28/astribot-corporations-s1-robot-promises-swift-and-skilled-assistance/</ref> AstriBot Corporation started operation in 2022 and took just a year to develop its first humanoid robot, S1.<ref>https://www.msn.com/en-us/news/other/china-s-s1-robot-impresses-with-its-human-like-speed-and-precision/ar-AA1nJ0BG</ref> Now, let's add this information to the article along with an infobox. == Overview == AstriBot Corporation, a subsidiary of Stardust Intelligence, is a Chinese tech company responsible for the creation and development of the Astribot S1, an efficient and capable humanoid robotic assistant. Known for its impressive parameters, such as the ability to lift up to 10 kilograms and move at speeds of 10 meters per second, Astribot S1 represents significant advancements in the field of AI-controlled humanoid robotics. The company began operations in 2022 and managed to develop its first humanoid robot, Astribot S1, in just a year. {{infobox company | name = AstriBot Corporation | country = China | website_link = | robots = Astribot S1 }} [[Category:Companies]] [[Category:Robotics]] == References == <references /> 55ce9381269bfe75632c051a5df02988fd482821 620 618 2024-04-29T00:41:33Z 194.195.93.78 0 wikitext text/x-wiki The [[Astribot S1]] is a product of the Chinese tech company Astribot Corporation, a subsidiary of Stardust Intelligence. This firm has made notable strides in the field of AI-controlled humanoid robots, developing the Astribot S1 as a highly efficient helper capable of lifting objects weighing up to 10 kilograms and moving at a speed of 10 meters per second.<ref>https://elblog.pl/2024/04/28/astribot-corporations-s1-robot-promises-swift-and-skilled-assistance/</ref> AstriBot Corporation started operation in 2022 and took just a year to develop its first humanoid robot, S1.<ref>https://www.msn.com/en-us/news/other/china-s-s1-robot-impresses-with-its-human-like-speed-and-precision/ar-AA1nJ0BG</ref> Now, let's add this information to the article along with an infobox. == Overview == AstriBot Corporation, a subsidiary of Stardust Intelligence, is a Chinese tech company responsible for the creation and development of the Astribot S1, an efficient and capable humanoid robotic assistant. Known for its impressive parameters, such as the ability to lift up to 10 kilograms and move at speeds of 10 meters per second, Astribot S1 represents significant advancements in the field of AI-controlled humanoid robotics. The company began operations in 2022 and managed to develop its first humanoid robot, Astribot S1, in just a year. {{infobox company | name = AstriBot Corporation | country = China | website_link = | robots = Astribot S1 }} [[Category:Companies]] [[Category:Robotics]] == References == <references /> e4b6722330d959ecf676de42ab69cd2a2599cab5 621 620 2024-04-29T00:41:41Z 194.195.93.78 0 wikitext text/x-wiki The [[Astribot S1]] is a product of the Chinese tech company Astribot Corporation, a subsidiary of Stardust Intelligence. This firm has made notable strides in the field of AI-controlled humanoid robots, developing the Astribot S1 as a highly efficient helper capable of lifting objects weighing up to 10 kilograms and moving at a speed of 10 meters per second.<ref>https://elblog.pl/2024/04/28/astribot-corporations-s1-robot-promises-swift-and-skilled-assistance/</ref> AstriBot Corporation started operation in 2022 and took just a year to develop its first humanoid robot, S1.<ref>https://www.msn.com/en-us/news/other/china-s-s1-robot-impresses-with-its-human-like-speed-and-precision/ar-AA1nJ0BG</ref> Now, let's add this information to the article along with an infobox. == Overview == AstriBot Corporation, a subsidiary of Stardust Intelligence, is a Chinese tech company responsible for the creation and development of the Astribot S1, an efficient and capable humanoid robotic assistant. Known for its impressive parameters, such as the ability to lift up to 10 kilograms and move at speeds of 10 meters per second, Astribot S1 represents significant advancements in the field of AI-controlled humanoid robotics. The company began operations in 2022 and managed to develop its first humanoid robot, Astribot S1, in just a year. {{infobox company | name = AstriBot Corporation | country = China | website_link = | robots = Astribot S1 }} == References == <references /> [[Category:Companies]] 61f63fa7bb147f695b9d8109eea1d54387c153bc 0 22 622 527 2024-04-29T00:42:07Z 194.195.93.78 0 wikitext text/x-wiki [[File:Optimus Tesla (1).jpg|right|200px|thumb]] Optimus is a humanoid robot developed by [[Tesla]], an American electric vehicle and clean energy company. Also known as Tesla Bot, Optimus is a key component of Tesla's expansion into automation and artificial intelligence technologies. {{infobox robot | name = Optimus | organization = [[Tesla]] | height = 5 ft 8 in (173 cm) | weight = 58 kg | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} == Development == Tesla initiated the development of the Optimus robot in 2021, with the goal of creating a multipurpose utility robot capable of performing unsafe, repetitive, or boring tasks primarily intended for a factory setting. Tesla's CEO, Elon Musk, outlined that Optimus could potentially transition into performing tasks in domestic environments in the future. == Design == The robot stands at a height of 5 feet 8 inches and weighs approximately 58 kilograms. Its design focusses on replacing human labor in hazardous environments, incorporating advanced sensors and algorithms to navigate complex workspaces safely. == Features == The features of Optimus are built around its capability to handle tools, carry out tasks requiring fine motor skills, and interact safely with human environments. The robot is equipped with Tesla's proprietary Full Self-Driving (FSD) computer, allowing it to understand and navigate real-world scenarios effectively. == Impact == Optimus has significant potential implications for labor markets, particularly in industries reliant on manual labor. Its development also sparks discussions on ethics and the future role of robotics in society. == References == * [https://www.tesla.com Tesla official website] * [https://www.youtube.com/watch?v=cpraXaw7dyc Presentation of Optimus by Tesla] [[Category:Robots]] 8ba29aa425ac39f70f5cad0bc2c5c560dfa3e217 0 126 623 599 2024-04-29T00:42:19Z 194.195.93.78 0 wikitext text/x-wiki CyberOne is a humanoid robot developed by the Chinese consumer electronics giant, Xiaomi. Unveiled in 2022 at a company event in Beijing by the founder, chairman, and CEO, Lei Jun, it is the newest member of Xiaomi's Cyber series, joining previously launched quadruped robots like CyberDog and CyberDog 2<ref>https://robotsguide.com/robots/cyberone</ref>. {{infobox robot | name = CyberOne | organization = [[Xiaomi]] | height = 177 cm | weight = 52 kg | single_arm_payload = 1.5 | runtime = | walk_speed = 3.6 km/h | video_link = https://www.youtube.com/watch?v=yBmatGQ0giY | cost = 600,000 - 700,000 yuan (est.) }} == Specifications == This bipedal humanoid robot has a height of 177 cm and weight of 52 kg, with an arm span of 168 cm<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. One of its distinct features is its ability to perceive 3D, recognize individuals, and respond to human emotions<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. Furthermore, it boasts a top speed of 3.6 km/ph<ref>https://www.theverge.com/2022/8/16/23307808/xiaomi-cyberone-humanoid-robot-tesla-optimus-bot-specs-comparison</ref>. == Pricing == The cost of CyberOne, if ever produced and made available for purchase, is estimated to be around 600,000 to 700,000 yuan<ref>https://robbreport.com/gear/electronics/xiaomi-humanoid-robot-cyberone-1234738597/</ref>. [[Category:Robots]] == References == <references /> 78fcd0eab316db9aa747f8a9e59cf1d0103c2fa6 0 138 624 603 2024-04-29T00:42:24Z 194.195.93.78 0 wikitext text/x-wiki Toyota Research Institute (TRI) is a subsidiary of the multinational automotive manufacturer, Toyota, and is based in Japan. The institute is exploring various facets of robotics and is responsible for building Toyota's third generation humanoid robot, the [[T-HR3]]<ref>[https://spectrum.ieee.org/toyota-gets-back-into-humanoid-robots-with-new-thr3 Toyota Gets Back Into Humanoid Robots With New T-HR3]</ref>. {{infobox company | name = Toyota Research Institute | country = Japan | website_link = https://global.toyota/ | robots = [[T-HR3]] }} == The T-HR3 Robot == The T-HR3 is a 1.5-meter humanoid robot designed to interact safely and naturally with its surroundings, and with the person controlling it<ref>[https://mag.toyota.co.uk/t-hr3-toyota-remote-control-robot/ T-HR3 - Toyota's new remote-controlled robot]</ref>. This robot weighs 75 kilograms and features 32 degrees of torque-controlled motion<ref>[https://spectrum.ieee.org/toyota-gets-back-into-humanoid-robots-with-new-thr3 Toyota Gets Back Into Humanoid Robots With New T-HR3]</ref>. It possesses some core capabilities: flexible joint control, whole-body coordination and balance control, and real remote maneuvering – thanks to its Torque Servo Module. These features enable the robot to control the force of contact it makes with any individual or object in its environment, to retain balance even when it collides with objects, and to mirror user movements with seamless and intuitive control<ref>[https://mag.toyota.co.uk/t-hr3-toyota-remote-control-robot/ T-HR3 - Toyota's new remote-controlled robot]</ref>. The development of the T-HR3 is championed by a research and development group led by Tomohisa Moridaira, based at Toyota's Tokyo Head Office.<ref>[https://global.toyota/en/newsroom/corporate/30609642.html Why is Toyota Developing Humanoid Robots?]</ref> == References == <references /> [[Category:Companies]] 9c3befc469e7cd2f5593df594577954052f5c309 0 111 625 594 2024-04-29T00:42:48Z 194.195.93.78 0 wikitext text/x-wiki The PX5 is a humanoid robot developed by Xpeng, unveiled for the first time during Xpeng Motors' Tech Day in 2023.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> The robot stands approximately 1.5 meters in height and is able to navigate different terrain and handle objects with precision, demonstrating remarkable stability.<ref>https://technode.com/2023/10/25/xpeng-tech-day-2023-first-mpv-self-driving-timeline-flying-cars-and-humanoid-robots/</ref> Constructed with a silver-white color scheme, the PX5 is also resistant to shock.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> {{infobox robot | name = PX5 | organization = [[Xpeng]] | height = 1.5 meters | weight = | video_link = https://www.youtube.com/watch?v=BNSZ8Fwcd20 | cost = }} == Development == Xpeng Robotics, an ecosystem company of Xpeng, specializing in smart robots, revealed the PX5. The company, which was founded in 2016, innovates in areas like robot powertrain, locomotion control, robot autonomy, robot interaction, and artificial intelligence contributing to a shared mission of exploring future mobility solutions.<ref>https://www.pxing.com/en/about</ref> == Design and Capabilities == The PX5 bears a striking silver-white finish and exhibits resistance to shock. Its capability to navigate through different terrains and handle handheld objects, such as a pen, with exceptional stability has been highlighted in demonstrations.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> <ref>https://technode.com/2023/10/25/xpeng-tech-day-2023-first-mpv-self-driving-timeline-flying-cars-and-humanoid-robots/</ref> == References == <references /> [[Category:Robots]] 2e0b3a4cde3f99bf769a1430f4ca8e1e7473db9b 0 7 626 505 2024-04-29T00:43:05Z 194.195.93.78 0 wikitext text/x-wiki Tesla, an innovative company known for its electric vehicles and sustainable energy solutions, has embarked on the development of a humanoid robot named Optimus. Optimus is envisioned to be a groundbreaking humanoid robot designed by Tesla to perform a wide range of tasks autonomously. Tesla's foray into humanoid robotics with Optimus showcases the company's commitment to advancing technology beyond the realm of electric vehicles. This project signifies Tesla's exploration of robotics and artificial intelligence beyond its traditional automotive focus. By venturing into humanoid robot development, Tesla aims to revolutionize industries and everyday tasks with cutting-edge automation. For more information about Tesla and its humanoid robot Optimus, interested individuals can visit Tesla's official website. {{infobox company | name = Tesla | country = United States | website_link = https://www.tesla.com/ | robots = [[Optimus]] }} [[Category: Companies]] 8f3e3f148f5bf53217a135dfa81954f14279d838 0 34 628 340 2024-04-29T01:52:40Z 194.195.93.80 0 wikitext text/x-wiki == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] c82f1e0fe0e614e045c68823bb1be532e1623391 0 153 633 2024-04-29T17:41:09Z 185.187.168.105 0 Created page with "Kayra is an open-source 3D printable robot. The project is avalable [https://github.com/assadollahi/kayra here]. [[Category: Robots]] [[Category: Stompy, Expand!]]" wikitext text/x-wiki Kayra is an open-source 3D printable robot. The project is avalable [https://github.com/assadollahi/kayra here]. [[Category: Robots]] [[Category: Stompy, Expand!]] 7115b976f10a17e4956934704886a012415e95f2 635 633 2024-04-29T18:04:19Z Stompy 14 Bot expanded article wikitext text/x-wiki '''Kayra''' is an open-source 3D printable humanoid robot designed for modify-ability and easy assembly. The project files, including design files for freeCAD, can be found on the [https://github.com/assadollahi/kayra GitHub platform]. Kayra's development is community-based, with goals to evolve the design in various directions.<ref>[https://github.com/assadollahi/kayra GitHub - assadollahi/kayra]</ref> Kayra is designed to be 3D printed and assembled with minimal effort. Ideally, no soldering is necessary for assembly. The software for Kayra can be written in Python for the host and Micropython for the controllers. It also includes simulation software, enabling users to work on the project even without physical hardware. The cost for a minimal system is intended to be low, enhancing accessibility and encouraging community participation.<ref>[https://github.com/assadollahi/kayra GitHub - assadollahi/kayra]</ref> Tutorial videos are available, providing real-time assembly instructions. For example, one recorded tutorial guides users through the process of assembling Kayra's left leg from printed parts, servos, bearings, and screws. The leg assembly can be completed within approximately 20 minutes.<ref>[https://kayra.org/ Home - the 3D printable, easy to modify, open-source humanoid]</ref> Please refer to the official [https://kayra.org/ Kayra website] and its [https://github.com/assadollahi/kayra GitHub page] for full details, shopping list, print instructions, and more. {{infobox robot | name = Kayra | organization = Open Source Community | video_link = | cost = Low | height = | weight = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = Active development }} [[Category: Robots]] [[Category: Open Source Robots]] [[Category: 3D Printable Robots]] == References == <references /> 5cabe363b1a418b4f1a230c9f58d76f3fc3c17d2 655 635 2024-04-29T19:28:51Z Ben 2 wikitext text/x-wiki '''Kayra''' is an open-source 3D printable humanoid robot designed for modify-ability and easy assembly. The project files, including design files for freeCAD, can be found on the [https://github.com/assadollahi/kayra GitHub platform]. Kayra's development is community-based, with goals to evolve the design in various directions.<ref>[https://github.com/assadollahi/kayra GitHub - assadollahi/kayra]</ref> Kayra is designed to be 3D printed and assembled with minimal effort. Ideally, no soldering is necessary for assembly. The software for Kayra can be written in Python for the host and Micropython for the controllers. It also includes simulation software, enabling users to work on the project even without physical hardware. The cost for a minimal system is intended to be low, enhancing accessibility and encouraging community participation.<ref>[https://github.com/assadollahi/kayra GitHub - assadollahi/kayra]</ref> Tutorial videos are available, providing real-time assembly instructions. For example, one recorded tutorial guides users through the process of assembling Kayra's left leg from printed parts, servos, bearings, and screws. The leg assembly can be completed within approximately 20 minutes.<ref>[https://kayra.org/ Home - the 3D printable, easy to modify, open-source humanoid]</ref> Please refer to the official [https://kayra.org/ Kayra website] and its [https://github.com/assadollahi/kayra GitHub page] for full details, shopping list, print instructions, and more. {{infobox robot | name = Kayra | organization = Open Source Community | video_link = | cost = Low | height = | weight = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = Active development }} == References == <references /> [[Category:Robots]] [[Category:Open Source]] 4600808d10da0493e037db3548ca5899dca72d80 0 41 636 204 2024-04-29T18:14:45Z Ben 2 wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Resources == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] 4e1c981e24ebe49198f05f433fbf564c8fd0164d 641 636 2024-04-29T18:20:55Z Ben 2 wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] f213abd306b6687a1a042e9379c751b442529baa 644 641 2024-04-29T19:05:45Z Matt 16 Add initial atopile information and ordering information wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] e45da6806676211370e2ff8f06ffd083ebf26f70 651 644 2024-04-29T19:27:02Z Ben 2 wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] dd02e3b1bcc54906783d0bea890acc2e5e774893 0 154 637 2024-04-29T18:15:21Z Ben 2 Created page with "[https://atopile.io/ atopile] is a language and toolchain to describe electronic circuit boards with code. [[Category: Stompy, Expand!]]" wikitext text/x-wiki [https://atopile.io/ atopile] is a language and toolchain to describe electronic circuit boards with code. [[Category: Stompy, Expand!]] 4e88bf0667de28907e1ccbf8e2eff9e479c6af0c 639 637 2024-04-29T18:19:36Z Ben 2 wikitext text/x-wiki [https://atopile.io/ atopile] is a language and toolchain to describe electronic circuit boards with code. {{#related:Building a PCB}} [[Category: Stompy, Expand!]] [[Category: Electronics]] d113d7c160db72cfd7c950b81597ecc76b176014 640 639 2024-04-29T18:20:43Z Ben 2 wikitext text/x-wiki [https://atopile.io/ atopile] is a language and toolchain to describe electronic circuit boards with code. === Related Articles === * [[Building a PCB]] [[Category: Stompy, Expand!]] [[Category: Electronics]] 344f0cd074e28ef476682122fc310f706b67b2ba 642 640 2024-04-29T19:04:16Z Stompy 14 Bot expanded article wikitext text/x-wiki Here is an expanded article on Atopile. Atopile is a programming language and integrated toolchain specifically designed for the process of describing electronic circuit boards, streamlining creation, and modification. Atopile provides an alternative approach to traditional schematic capture methods, offering a way to transcribe circuit design into organized, scalable, and reusable code. {{infobox company | name = Atopile | country = [Unknown] | website_link = https://atopile.io/ | robots = [None] }} == Overview == Atopile's unique approach of treating hardware design like software development paves the way for features such as version control, automated testing, and modularity, which have long been staples in the software development industry. By allowing circuit designs to be written, understood, and manipulated as code, Atopile enables a new level of flexibility and efficiency for electronics engineers. == Functionality == The Atopile language and toolchain is equipped with a range of features that enhance both the design and prototyping stages of electronic circuit board development. Some core functionalities include: * '''Iterative Development:''' Atopile supports version control, allowing for iterative development, history tracking, and collaborative capabilities, similar to software development environments. This process avails design revision and controlled evolution of circuit boards, improving product quality over time because earlier designs can be updated and improved rather than discarded. * '''Modularity:''' With the language used in Atopile, circuit designs can be broken down into modules, making them reusable across different projects. This function promotes better code organization and reduces redundancy. * '''Automated Testing:''' Like in software development, automated testing can be conducted within Atopile's ecosystem, ensuring correctness and reliability of designs and minimizing chances of hardware failures. * '''Compatibility:''' Atopile's code can be exported to compatible formats for other electronic design automation (EDA) tools. This offers a higher level of versatility and interoperability with other development environments. == Use Cases == Atopile can be utilized for a wide range of applications, including academic research, commercial product development, prototyping, DIY electronics project, and educational purposes. == Constructive Criticism == As with any tool, Atopile is not without its criticisms. One primary concern is the steep learning curve associated with mastering a new language. This might pose difficulties for those not familiar with software development principles. Furthermore, while the abstraction of circuit design into code offers many benefits, it may conceal certain hardware complexities that are more intuitively understood in traditional visual schematic form. == References == <references /> === Related Articles === * [[Building a PCB]] [[Category: Electronics]] 8880bfe03ef6f0200ad229c9f0db1af69642ae4f 645 642 2024-04-29T19:06:51Z Ben 2 wikitext text/x-wiki Atopile is a programming language and integrated toolchain specifically designed for the process of describing electronic circuit boards, streamlining creation, and modification. Atopile provides an alternative approach to traditional schematic capture methods, offering a way to transcribe circuit design into organized, scalable, and reusable code. {{infobox company | name = Atopile | country = United States | website_link = https://atopile.io/ }} == Overview == Atopile's unique approach of treating hardware design like software development paves the way for features such as version control, automated testing, and modularity, which have long been staples in the software development industry. By allowing circuit designs to be written, understood, and manipulated as code, Atopile enables a new level of flexibility and efficiency for electronics engineers. == Functionality == The Atopile language and toolchain is equipped with a range of features that enhance both the design and prototyping stages of electronic circuit board development. Some core functionalities include: * '''Iterative Development:''' Atopile supports version control, allowing for iterative development, history tracking, and collaborative capabilities, similar to software development environments. This process avails design revision and controlled evolution of circuit boards, improving product quality over time because earlier designs can be updated and improved rather than discarded. * '''Modularity:''' With the language used in Atopile, circuit designs can be broken down into modules, making them reusable across different projects. This function promotes better code organization and reduces redundancy. * '''Automated Testing:''' Like in software development, automated testing can be conducted within Atopile's ecosystem, ensuring correctness and reliability of designs and minimizing chances of hardware failures. * '''Compatibility:''' Atopile's code can be exported to compatible formats for other electronic design automation (EDA) tools. This offers a higher level of versatility and interoperability with other development environments. == Use Cases == Atopile can be utilized for a wide range of applications, including academic research, commercial product development, prototyping, DIY electronics project, and educational purposes. == Related Articles == * [[Building a PCB]] == References == <references /> [[Category: Electronics]] 19c82d4e718a9555656ac519604bd09f344e31c3 0 155 638 2024-04-29T18:17:09Z Ben 2 Created page with "The Controller Area Network is a vehicle bus standard designed to allow microcontrollers and devices to communicate with each other. === MCP2515 === The MCP2515 chip is an i..." wikitext text/x-wiki The Controller Area Network is a vehicle bus standard designed to allow microcontrollers and devices to communicate with each other. === MCP2515 === The MCP2515 chip is an integrated circuit that acts as a stand-alone CAN controller with SPI interface. [[Category: Stompy, Expand!]] a7d5f16821b9db1493f1d7b3608f39545120afb7 643 638 2024-04-29T19:04:46Z Stompy 14 Bot expanded article wikitext text/x-wiki {{infobox actuator | name = MCP2515 | manufacturer = Microchip Technology | cost = Not listed | purchase_link = Not listed | nominal_torque = Not applicable | peak_torque = Not applicable | weight = Not applicable | dimensions = Not listed | gear_ratio = Not applicable | voltage = Not listed | cad_link = Not listed | interface = SPI interface | gear_type = Not applicable }} == Controller Area Network == The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == References == <references /> [[Category: Controller Area Network]] [[Category: Integrated Circuits]] [[Category: Automotive Technology]] [[Category: Robotics Communications]] 9935ca0e82d1837ac65a1dfdd4d394afbb60023f 646 643 2024-04-29T19:07:18Z Ben 2 wikitext text/x-wiki {{infobox actuator | name = MCP2515 | manufacturer = Microchip Technology | cost = Not listed | purchase_link = Not listed | nominal_torque = Not applicable | peak_torque = Not applicable | weight = Not applicable | dimensions = Not listed | gear_ratio = Not applicable | voltage = Not listed | cad_link = Not listed | interface = SPI interface | gear_type = Not applicable }} == Controller Area Network == The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == References == <references /> 0e7169ff1fa8c99f2c6bbbd801a3e0d8739250d6 647 646 2024-04-29T19:07:32Z Ben 2 wikitext text/x-wiki == Controller Area Network == The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == References == <references /> 39616c557f52b6113b96a1dd89c5748ae513ec67 648 647 2024-04-29T19:07:40Z Ben 2 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == References == <references /> 614e870bceeec16427c679eed2852cbb10cf2edb 0 156 649 2024-04-29T19:24:50Z Ben 2 Created page with "This document acts as a progress tracker for the [[Stompy]] robot build. == To Do == === Teleop === * Create a VR interface === Firmware === * Get * === Electronics ===..." wikitext text/x-wiki This document acts as a progress tracker for the [[Stompy]] robot build. == To Do == === Teleop === * Create a VR interface === Firmware === * Get * === Electronics === * Understand how to build a PCB and update the [[Building a PCB]] guide ** Get PCBs designed and shipped ** Two PCBs: One for the head and one for the body *** Head PCB will mount a Jetson Nano. It will have two MIPI CSI ports for two cameras from the eyes, plus audio and speakers. The Jetson Nano will act as a "tokenizer" which multiplexes the signals over ethernet to the main board in the body *** The body PCB will mount a full Jetson Orin, as well as the IMU, CAN tranceivers, and power management * Debug issues with the [[CAN Bus]] - specifically, write our own driver for the MCP2515 IC (one of the most common ICs for sending and receiving messages over a CAN bus) 9a09320fe88f3c53fd3dc5655abf9af384bcc05d 650 649 2024-04-29T19:25:37Z Ben 2 wikitext text/x-wiki This document acts as a progress tracker for the [[Stompy]] robot build. == To Do == === Teleop === * Create a VR interface for controlling Stompy in simulation * Port VR interface to work on the full robot === Firmware === * Debug issues with the [[CAN Bus]] - specifically, write our own driver for the MCP2515 IC (one of the most common ICs for sending and receiving messages over a CAN bus) * Set up continuous integration for building and flashing the full operating system === Electronics === * Understand how to build a PCB and update the [[Building a PCB]] guide ** Get PCBs designed and shipped ** Two PCBs: One for the head and one for the body *** Head PCB will mount a Jetson Nano. It will have two MIPI CSI ports for two cameras from the eyes, plus audio and speakers. The Jetson Nano will act as a "tokenizer" which multiplexes the signals over ethernet to the main board in the body *** The body PCB will mount a full Jetson Orin, as well as the IMU, CAN tranceivers, and power management 2b51e0576fc0d8cedc6a549f38b4dfffb590264b 661 650 2024-04-29T19:31:26Z Ben 2 wikitext text/x-wiki This document acts as a progress tracker for the [[Stompy]] robot build. == To Do == === Teleop === * Create a VR interface for controlling Stompy in simulation * Port VR interface to work on the full robot === Firmware === * Debug issues with the [[CAN Bus]] - specifically, write our own driver for the MCP2515 IC (one of the most common ICs for sending and receiving messages over a CAN bus) * Set up continuous integration for building and flashing the full operating system === Actuator === * Build on top of a good open-source actuator like the [[SPIN Drive]] or [[OBot]] actuator to design and fabricate our own open-source actuator === Electronics === * Understand how to build a PCB and update the [[Building a PCB]] guide ** Get PCBs designed and shipped ** Two PCBs: One for the head and one for the body *** Head PCB will mount a Jetson Nano. It will have two MIPI CSI ports for two cameras from the eyes, plus audio and speakers. The Jetson Nano will act as a "tokenizer" which multiplexes the signals over ethernet to the main board in the body *** The body PCB will mount a full Jetson Orin, as well as the IMU, CAN tranceivers, and power management 79ce754f2b12deb624c8aa0bb78062287cd2b8eb 662 661 2024-04-29T19:31:37Z Ben 2 wikitext text/x-wiki This document acts as a progress tracker for the [[Stompy]] robot build. == To Do == === Teleop === * Create a VR interface for controlling Stompy in simulation * Port VR interface to work on the full robot === Firmware === * Debug issues with the [[CAN Bus]] - specifically, write our own driver for the MCP2515 IC (one of the most common ICs for sending and receiving messages over a CAN bus) * Set up continuous integration for building and flashing the full operating system === Actuator === * Build on top of a good open-source actuator like the [[SPIN Servo]] or [[OBot]] actuator to design and fabricate our own open-source actuator === Electronics === * Understand how to build a PCB and update the [[Building a PCB]] guide ** Get PCBs designed and shipped ** Two PCBs: One for the head and one for the body *** Head PCB will mount a Jetson Nano. It will have two MIPI CSI ports for two cameras from the eyes, plus audio and speakers. The Jetson Nano will act as a "tokenizer" which multiplexes the signals over ethernet to the main board in the body *** The body PCB will mount a full Jetson Orin, as well as the IMU, CAN tranceivers, and power management 59fb9df3f41b2096a50905959e9fa55647f5c197 663 662 2024-04-29T19:33:58Z Ben 2 wikitext text/x-wiki This document acts as a progress tracker for the [[Stompy]] robot build. == To Do == === Teleop === * Create a VR interface for controlling Stompy in simulation * Port VR interface to work on the full robot === Firmware === * Debug issues with the [[CAN Bus]] - specifically, write our own driver for the MCP2515 IC (one of the most common ICs for sending and receiving messages over a CAN bus) * Set up continuous integration for building and flashing the full operating system === Actuator === * Build on top of a good open-source actuator like the [[SPIN Servo]], [[MIT Cheetah]] or [[OBot]] actuator to design and fabricate our own open-source actuator === Electronics === * Understand how to build a PCB and update the [[Building a PCB]] guide ** Get PCBs designed and shipped ** Two PCBs: One for the head and one for the body *** Head PCB will mount a Jetson Nano. It will have two MIPI CSI ports for two cameras from the eyes, plus audio and speakers. The Jetson Nano will act as a "tokenizer" which multiplexes the signals over ethernet to the main board in the body *** The body PCB will mount a full Jetson Orin, as well as the IMU, CAN tranceivers, and power management c441f8a9aae494590ecc65bf4d11642640e16fed 0 90 652 566 2024-04-29T19:27:47Z Ben 2 wikitext text/x-wiki {{infobox actuator | name = SPIN Servo | manufacturer = Holry Motor | cost = USD 30 (BOM) | purchase_link = https://shop.atopile.io/ | nominal_torque = 0.125nm | peak_torque = 0.375nm | weight = 311.6g | dimensions = 42mmx42mmx60mm | gear_ratio = Direct drive (bolt on options) | voltage = 12V-24V | cad_link = https://github.com/atopile/spin-servo-drive/tree/main/mech | interface = CAN bus, i2c }} == Overview == The SPIN Servo is an open-source hardware project, designed to make the use of fully-fledged Brushless DC (BLDC) servo motors easy and cost-effective<ref>[https://github.com/atopile/spin-servo-drive GitHub - atopile/spin-servo-drive: SPIN - Servos are awesome]</ref>. It is primarily engineered by atopile, which is known for its toolchains to describe electronic circuit boards with code<ref>[https://atopile.io/spin/ SPIN - atopile]</ref>. The intention behind this project is to introduce software development workflows like reuse, validation, and automation into the world of electronics. The SPIN Servo is manufactured by the Holry Motor company. It weighs 311.6g and its dimensions are 42mm x 42mm x 60mm. The cost of the Bill of Materials (BOM) is USD 30. The nominal torque of the spin servo is 0.125nm, while it can peak at 0.375nm. The interface employed by this servo is CAN bus and i2c. It operates at a voltage of 12V-24V and its gear ratio is directly driven, although there are bolt-on options available. All designs and schematics related to the SPIN Servo can be found at their official GitHub repository<ref>[https://github.com/atopile/spin-servo-drive/tree/main/mech SPIN Servo CAD - GitHub]</ref>. Interested individuals can purchase the SPIN Servo from the official atopile shop<ref>[https://shop.atopile.io/ Atopile Shop]</ref>. == References == <references /> [[Category:Actuators]] [[Category:Open Source]] 23a86b767871a4cecb7f4401a2d6fcf758be62da 14 157 653 2024-04-29T19:28:17Z Ben 2 Created page with "This is a tag for projects which are open-source, meaning that you can see the full source code or reference design online." wikitext text/x-wiki This is a tag for projects which are open-source, meaning that you can see the full source code or reference design online. 4e1aa0a75a7c9435ba02eb3faa92d44ac44a282c 0 89 654 380 2024-04-29T19:28:29Z Ben 2 wikitext text/x-wiki The [https://github.com/unhuman-io/obot OBot] is an open-source robot designed for doing mobile manipulation. {{infobox actuator | name = OBot | voltage = 36V | interface = USB }} [[Category:Actuators]] [[Category:Open Source]] 30ae1a1677c60e74e065e7c3b42eeaf2930021c4 0 2 656 258 2024-04-29T19:29:12Z Ben 2 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. See the [[Stompy Build Guide|build guide]] for a walk-through of how to build one yourself. = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] [[Category:Open Source]] e8361cf4bdf128021548ec040d15f56e3818d149 657 656 2024-04-29T19:29:40Z Ben 2 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some additional relevant links: - [[Stompy To-Do List]] - [[Stompy Build Guide]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] [[Category:Open Source]] 01d49b67e3b9bea5c5dbc1342b46867c66e1229f 658 657 2024-04-29T19:29:49Z Ben 2 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some additional relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] [[Category:Open Source]] 7f573535d974c1f16f164e4da7f0899ee39341c4 659 658 2024-04-29T19:30:19Z Ben 2 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] [[Category:Open Source]] 33c44a58c07be35e2c900baebd97148d361b59e1 660 659 2024-04-29T19:30:31Z Ben 2 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] [[Category:Open Source]] 48c3a6cc32ac0e37d5c2a41f9f10e003ca621c39 0 158 664 2024-04-29T19:34:08Z Ben 2 Created page with "The '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. [[Category:Actuators]] [[Category:Open Source]]" wikitext text/x-wiki The '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. [[Category:Actuators]] [[Category:Open Source]] 6acfa8a494e4e34bd08d0fd53caae78b9a02b33b 0 158 665 664 2024-04-29T19:34:37Z Ben 2 wikitext text/x-wiki The '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. [[Category:Actuators]] [[Category:Open Source]] [[Category:Stompy, Expand!]] 606f8209b269ef63917f9c250b46129808ca17fc 666 665 2024-04-29T20:04:32Z Stompy 14 Bot expanded article wikitext text/x-wiki '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. This robot is developed in the lab of Professor Sangbae Kim at MIT<ref>https://www.csail.mit.edu/news/one-giant-leap-mini-cheetah</ref>. The MIT Cheetah is known for its agility and ability to adapt to varying terrain conditions without requiring a terrain map in advance<ref>https://www.csail.mit.edu/news/one-giant-leap-mini-cheetah</ref>. {{infobox robot | name = MIT Cheetah | organization = MIT | video_link = | cost = | height = | weight = 20 pounds | speed = Twice the speed of average human walking speed<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>,<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref> | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = Active }} == Design and Development == The MIT Cheetah displays remarkable adaptability in its functions. Despite being only 20 pounds, the robot can bend and swing its legs wide, allowing it to either walk right side up or upside down<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>. This flexibility is a result of its design that prioritizes a wide range of motion. The robot is further equipped to deal with challenges posed by rough, uneven terrain. It can traverse such landscapes at a pace twice as fast as an average person's walking speed<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref>. The MIT Cheetah’s design, particularly the implementation of a control system that enables agile running, has been largely driven by the "learn-by-experience model". This approach, in contrast to previous designs reliant primarily on human analytical insights, allows the robot to respond quickly to changes in the environment<ref>https://news.mit.edu/2022/3-questions-how-mit-mini-cheetah-learns-run-fast-0317</ref>. == See also == * [[Sangbae Kim]] * [[Low-inertia actuator]] == References == <references /> [[Category:Actuators]] [[Category:Open Source]] [[Category:Robots]] 3d3130a1540c5784fc19f305d84994a8c68900a3 667 666 2024-04-29T20:07:43Z Ben 2 wikitext text/x-wiki '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. This robot is developed in the lab of Professor Sangbae Kim at MIT<ref>https://www.csail.mit.edu/news/one-giant-leap-mini-cheetah</ref>. The MIT Cheetah is known for its agility and ability to adapt to varying terrain conditions without requiring a terrain map in advance<ref>https://www.csail.mit.edu/news/one-giant-leap-mini-cheetah</ref>. {{infobox robot | name = MIT Cheetah | organization = MIT | video_link = | cost = | height = | weight = 20 pounds | speed = Twice the speed of average human walking speed<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>,<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref> | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = Active }} == Design and Development == The MIT Cheetah displays remarkable adaptability in its functions. Despite being only 20 pounds, the robot can bend and swing its legs wide, allowing it to either walk right side up or upside down<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>. This flexibility is a result of its design that prioritizes a wide range of motion. The robot is further equipped to deal with challenges posed by rough, uneven terrain. It can traverse such landscapes at a pace twice as fast as an average person's walking speed<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref>. The MIT Cheetah’s design, particularly the implementation of a control system that enables agile running, has been largely driven by the "learn-by-experience model". This approach, in contrast to previous designs reliant primarily on human analytical insights, allows the robot to respond quickly to changes in the environment<ref>https://news.mit.edu/2022/3-questions-how-mit-mini-cheetah-learns-run-fast-0317</ref>. == References == <references /> [[Category:Actuators]] [[Category:Open Source]] [[Category:Robots]] e07bb2c4d991044cf16c8b82d6cbb41c0dd2ab34 668 667 2024-04-29T20:08:16Z Ben 2 wikitext text/x-wiki '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. This robot is developed in the lab of Professor Sangbae Kim at MIT<ref>https://www.csail.mit.edu/news/one-giant-leap-mini-cheetah</ref>. The MIT Cheetah is known for its agility and ability to adapt to varying terrain conditions without requiring a terrain map in advance {{infobox robot | name = MIT Cheetah | organization = MIT | video_link = | cost = | height = | weight = 20 pounds | speed = Twice the speed of average human walking speed<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>,<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref> | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = Active }} == Design and Development == The MIT Cheetah displays remarkable adaptability in its functions. Despite being only 20 pounds, the robot can bend and swing its legs wide, allowing it to either walk right side up or upside down<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>. This flexibility is a result of its design that prioritizes a wide range of motion. The robot is further equipped to deal with challenges posed by rough, uneven terrain. It can traverse such landscapes at a pace twice as fast as an average person's walking speed<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref>. The MIT Cheetah’s design, particularly the implementation of a control system that enables agile running, has been largely driven by the "learn-by-experience model". This approach, in contrast to previous designs reliant primarily on human analytical insights, allows the robot to respond quickly to changes in the environment<ref>https://news.mit.edu/2022/3-questions-how-mit-mini-cheetah-learns-run-fast-0317</ref>. == References == <references /> [[Category:Actuators]] [[Category:Open Source]] [[Category:Robots]] b9b5d78c389c713ece94fd172b4de2d7a5c100c5 669 668 2024-04-29T20:26:22Z Ben 2 wikitext text/x-wiki '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. This robot is developed in the lab of Professor Sangbae Kim at MIT<ref>https://www.csail.mit.edu/news/one-giant-leap-mini-cheetah</ref>. The MIT Cheetah is known for its agility and ability to adapt to varying terrain conditions without requiring a terrain map in advance {{infobox robot | name = MIT Cheetah | organization = MIT | video_link = | cost = | height = | weight = 20 pounds | speed = Twice the speed of average human walking speed<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>,<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref> | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = Active }} == Design and Development == The MIT Cheetah displays remarkable adaptability in its functions. Despite being only 20 pounds, the robot can bend and swing its legs wide, allowing it to either walk right side up or upside down<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>. This flexibility is a result of its design that prioritizes a wide range of motion. The robot is further equipped to deal with challenges posed by rough, uneven terrain. It can traverse such landscapes at a pace twice as fast as an average person's walking speed<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref>. The MIT Cheetah’s design, particularly the implementation of a control system that enables agile running, has been largely driven by the "learn-by-experience model". This approach, in contrast to previous designs reliant primarily on human analytical insights, allows the robot to respond quickly to changes in the environment<ref>https://news.mit.edu/2022/3-questions-how-mit-mini-cheetah-learns-run-fast-0317</ref>. === Chips === * [https://www.st.com/en/microcontrollers-microprocessors/stm32-32-bit-arm-cortex-mcus.html STM32 32-buit Arm Cortex MCU] * [https://www.ti.com/product/DRV8323 DRV8323 3-phase smart gate driver] * [https://www.monolithicpower.com/en/ma702.html MA702 angular position measurement device] * [https://www.microchip.com/en-us/product/mcp2515 MCP2515 CAN Controller with SPI Interface] == References == <references /> [[Category:Actuators]] [[Category:Open Source]] [[Category:Robots]] 84ab0d43b891b17a6428395d5eff1d8f67d0e3a3 0 139 670 482 2024-04-29T20:54:09Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Onboarding === * Watch out for an email from Gusto (our HR software), with an official offer letter and instructions on how to onboard you into our system. * Once you accept, Ben will add you to the system, after which you will have to enter your bank account information in order to be paid. === Pre-Internship Checklist === * Create a wiki account and mark yourself as an employee (you can use [[User:Ben]] as a template). You'll use your account as the main way to keep track of what you've done over the course of the internship. * Contribute an article about something you find interesting. See the [[Contributing]] guide. === Additional Notes === * For travel expenses, please purchase your own flight and keep your receipts so that we can reimburse you later. 6d4cb77692aa636842375cf984f2117d4f3241d6 722 670 2024-04-30T05:06:10Z 108.211.178.220 0 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Onboarding === * Watch out for an email from Gusto (our HR software), with an official offer letter and instructions on how to onboard you into our system. * Once you accept, Ben will add you to the system, after which you will have to enter your bank account information in order to be paid. === Pre-Internship Checklist === * Create a wiki account and mark yourself as an employee (you can use [[User:Ben]] as a template). You'll use your account as the main way to keep track of what you've done over the course of the internship. * Contribute an article about something you find interesting. See the [[Contributing]] guide. === Additional Notes === * For travel expenses, please purchase your own flight and keep your receipts so that we can reimburse you later. [[Category:K-Scale]] 94c2c2fcb73cfa09cbf274788d6c39397ca7e444 0 1 671 634 2024-04-29T22:57:21Z 106.213.81.175 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 88b296ec12507a5a311563b5e7ab5b1a30a42022 695 671 2024-04-29T23:53:52Z Admin 1 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project to apply [[VESC]] to |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} d5370c303d29291fc299f9b33f4f882f16211bd9 696 695 2024-04-30T00:21:53Z Admin 1 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project to apply [[VESC]] to |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 889e376cb04b5e9bafa5b752aac64dcc508d7c1b 700 696 2024-04-30T00:33:41Z Admin 1 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 87c487d12821493c2fc5526045982ed0f6c8e51b 6 161 680 2024-04-29T23:23:18Z Ben 2 wikitext text/x-wiki RMD X8-H actuator PCB 19d8cf3fbc55385c7de6702ffba955fd23b0b153 0 155 685 648 2024-04-29T23:29:05Z Ben 2 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == References == <references /> [[Category:Communication]] 32519f6d3ebd9eec70468759b564eaf366c41136 686 685 2024-04-29T23:30:20Z Admin 1 Admin moved page [[CAN Bus]] to [[Controller Area Network (CAN)]]: Better name wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == References == <references /> [[Category:Communication]] 32519f6d3ebd9eec70468759b564eaf366c41136 0 162 687 2024-04-29T23:30:20Z Admin 1 Admin moved page [[CAN Bus]] to [[Controller Area Network (CAN)]]: Better name wikitext text/x-wiki #REDIRECT [[Controller Area Network (CAN)]] e273345fd34037f23609e48e5318803810e08bc4 6 163 688 2024-04-29T23:49:38Z Admin 1 wikitext text/x-wiki Vescular6 PCB 3d94c008d77600b7f4b57f60e895a2e76dad232c 0 164 689 2024-04-29T23:49:57Z Admin 1 Created page with "[https://dongilc.gitbook.io/openrobot-inc VESCular6] is an open-source motor controller based on [[VESC]]. [[File:Vescular6.png|thumb|left]]" wikitext text/x-wiki [https://dongilc.gitbook.io/openrobot-inc VESCular6] is an open-source motor controller based on [[VESC]]. [[File:Vescular6.png|thumb|left]] c43c591d0ebb2f882581cc2b0abf9a4cb3eef9e7 0 165 690 2024-04-29T23:50:52Z Admin 1 Created page with "The [https://vesc-project.com/vesc_tool VESC project] is an open-source [[ESC]] implementation<ref>https://vedder.se/2015/01/vesc-open-source-esc/</ref>. === References === <..." wikitext text/x-wiki The [https://vesc-project.com/vesc_tool VESC project] is an open-source [[ESC]] implementation<ref>https://vedder.se/2015/01/vesc-open-source-esc/</ref>. === References === <references/> f893322eb5042e1bd6a9c7b256d065123942668f 0 166 691 2024-04-29T23:51:41Z Admin 1 Created page with "An ```Electronic Speed Control``` is an electronic circuit that controls and regulates the speed of an electric motor." wikitext text/x-wiki An ```Electronic Speed Control``` is an electronic circuit that controls and regulates the speed of an electric motor. 6f03da7f302cd575dbfb3014276db83bc32b61e4 692 691 2024-04-29T23:51:51Z Admin 1 wikitext text/x-wiki An '''Electronic Speed Control''' is an electronic circuit that controls and regulates the speed of an electric motor. b0dc2949e65b0a8e26aa405e06cb4f55430411ba 693 692 2024-04-29T23:52:06Z Admin 1 Admin moved page [[ESC]] to [[Electronic Speed Control]] wikitext text/x-wiki An '''Electronic Speed Control''' is an electronic circuit that controls and regulates the speed of an electric motor. b0dc2949e65b0a8e26aa405e06cb4f55430411ba 0 167 694 2024-04-29T23:52:06Z Admin 1 Admin moved page [[ESC]] to [[Electronic Speed Control]] wikitext text/x-wiki #REDIRECT [[Electronic Speed Control]] 17a0b02bfb36300ffcdde3cc07bc66c2753ceebc 0 168 697 2024-04-30T00:22:22Z Admin 1 Created page with "'''EtherCAT (Ethernet for Control Automation Technology)''' is an Ethernet-based fieldbus system developed by Beckhoff Automation. [[Category:Stompy, Expand!]]" wikitext text/x-wiki '''EtherCAT (Ethernet for Control Automation Technology)''' is an Ethernet-based fieldbus system developed by Beckhoff Automation. [[Category:Stompy, Expand!]] 1982882bad0a3fe09db89fe3d0a5dc990d4f43d9 698 697 2024-04-30T00:31:38Z Admin 1 wikitext text/x-wiki '''EtherCAT (Ethernet for Control Automation Technology)''' is an Ethernet-based fieldbus system developed by Beckhoff Automation. [[Category:Stompy, Expand!]] [[Category:Communication]] 05a1905b64bb67eb3b3c710cbe122d2f04ace42f 703 698 2024-04-30T01:04:29Z Stompy 14 Bot expanded article wikitext text/x-wiki EtherCAT (Ethernet for Control Automation Technology) is a reliable, real-time Ethernet-based fieldbus system initially developed and implemented by Beckhoff Automation. Remarkably effective, this system uses a highly efficient communication method that allows data to be transferred from an industrial computer and communicated downstream to each piece of equipment in a daisy-chain configuration. The EtherCAT protocol is open and has gained wide acceptance due to its real-time capabilities and robust performance in industrial environments. The protocol was developed to offer high performance, real-time capabilities, and cost-effectiveness for systems that range from small to large scale. It finds a vast array of applications in machine controls, as well as in robotics and other fields that require high-speed, reliable communication between controllers and devices. This system is compatible with standard Ethernet and, therefore, does not require any specific hardware, which makes it a cost-effective solution. EtherCAT's ability to function on existing networks means it can work in tandem with other protocols and can share the same cables, switches, and the like, including industrial Ethernet infrastructure. == History == The EtherCAT technology was initially introduced by Beckhoff Automation in 2003. It was then smoothly transferred to the EtherCAT Technology Group (ETG) and has since been developed and maintained by this group. ETG is an open, non-profit organization that offers membership to manufacturers, developers, and users. The protocol has now established itself globally, becoming a high-performance Ethernet communication system in automation technology due to its real-time capabilities and the ability to connect a large number of devices to a single network. == Key Features of EtherCAT == EtherCAT offers several distinguishing features that contribute to its wide-spread use: * '''High Speed''': The system can process 1000 distributed I/O in 30 μs, or communicate with 100 servo axes in 100 μs. * '''Efficiency''': Data is not just passed along the communication line, but devices can directly read and write data as it flows by. * '''Robustness''': EtherCAT features error detection and flexible topology configurations to attain robustness against failure or data corruption. * '''Operability''': It allows for online configuration and diagnostics, making it a flexible and operable solution. * '''Cost-effectiveness''': Requires minimal hardware and can function over standard Ethernet, making its implementation cost-effective compared to other fieldbus systems. == Applications == The efficiency, robustness, and high speed of EtherCAT have made it a preferred choice in a host of applications: * '''Industrial Automation''': EtherCAT's high-speed, deterministic communication and robust error detection make it widely accepted in industrial automation, such as assembly lines, packing machinery, and material handling. * '''Robotics''': The perfect choice for complex servo control systems due to its capacity to handle fast control loops and synchronize large amounts of data. * '''Wind Energy''': EtherCAT is commonly used to control and monitor wind turbines due to its ability to accurately control multiple axes simultaneously. * '''Medical Technology''': In medical technology, EtherCAT is embraced for applications such as medical imaging, where it can handle the transfer of large amounts of data in real time. {{infobox protocol | name = EtherCAT | organization = Beckhoff Automation | introduced_date = 2003 | standard = IEC 61158, IEC 61784-2 | network_type = Ethernet-based | topology = Daisy-chain, star, tree, or mixed arrangements | speed = Up to 100 Mbit/s | distance = Up to 100m (over copper cable), Up to 20km (over fiber) | website_link = https://www.ethercat.org/ }} [[Category:Communication]] [[Category:Automation]] [[Category:Fieldbus systems]] [[Category:Industrial Ethernet]] [[Category:Robotics]] == References == <references /> [Assistant Used Tavily Search API with the query "EtherCAT Beckhoff Automation" to gather the relevant background, key features, applications, and other specifics about EtherCAT.] b2159ee143fb8e96a38600726c7794c4a605386e 725 703 2024-04-30T05:24:50Z 108.211.178.220 0 wikitext text/x-wiki EtherCAT (Ethernet for Control Automation Technology) is a reliable, real-time Ethernet-based fieldbus system initially developed and implemented by Beckhoff Automation. Remarkably effective, this system uses a highly efficient communication method that allows data to be transferred from an industrial computer and communicated downstream to each piece of equipment in a daisy-chain configuration. The EtherCAT protocol is open and has gained wide acceptance due to its real-time capabilities and robust performance in industrial environments. The protocol was developed to offer high performance, real-time capabilities, and cost-effectiveness for systems that range from small to large scale. It finds a vast array of applications in machine controls, as well as in robotics and other fields that require high-speed, reliable communication between controllers and devices. This system is compatible with standard Ethernet and, therefore, does not require any specific hardware, which makes it a cost-effective solution. EtherCAT's ability to function on existing networks means it can work in tandem with other protocols and can share the same cables, switches, and the like, including industrial Ethernet infrastructure. == History == The EtherCAT technology was initially introduced by Beckhoff Automation in 2003. It was then smoothly transferred to the EtherCAT Technology Group (ETG) and has since been developed and maintained by this group. ETG is an open, non-profit organization that offers membership to manufacturers, developers, and users. The protocol has now established itself globally, becoming a high-performance Ethernet communication system in automation technology due to its real-time capabilities and the ability to connect a large number of devices to a single network. == Key Features of EtherCAT == EtherCAT offers several distinguishing features that contribute to its wide-spread use: * '''High Speed''': The system can process 1000 distributed I/O in 30 μs, or communicate with 100 servo axes in 100 μs. * '''Efficiency''': Data is not just passed along the communication line, but devices can directly read and write data as it flows by. * '''Robustness''': EtherCAT features error detection and flexible topology configurations to attain robustness against failure or data corruption. * '''Operability''': It allows for online configuration and diagnostics, making it a flexible and operable solution. * '''Cost-effectiveness''': Requires minimal hardware and can function over standard Ethernet, making its implementation cost-effective compared to other fieldbus systems. == Applications == The efficiency, robustness, and high speed of EtherCAT have made it a preferred choice in a host of applications: * '''Industrial Automation''': EtherCAT's high-speed, deterministic communication and robust error detection make it widely accepted in industrial automation, such as assembly lines, packing machinery, and material handling. * '''Robotics''': The perfect choice for complex servo control systems due to its capacity to handle fast control loops and synchronize large amounts of data. * '''Wind Energy''': EtherCAT is commonly used to control and monitor wind turbines due to its ability to accurately control multiple axes simultaneously. * '''Medical Technology''': In medical technology, EtherCAT is embraced for applications such as medical imaging, where it can handle the transfer of large amounts of data in real time. {{infobox protocol | name = EtherCAT | organization = Beckhoff Automation | introduced_date = 2003 | standard = IEC 61158, IEC 61784-2 | network_type = Ethernet-based | topology = Daisy-chain, star, tree, or mixed arrangements | speed = Up to 100 Mbit/s | distance = Up to 100m (over copper cable), Up to 20km (over fiber) | website_link = https://www.ethercat.org/ }} [[Category:Communication]] == References == <references /> 77e60a24bbd337fc76793e3759ab4fc051782981 0 169 699 2024-04-30T00:32:54Z Admin 1 Created page with "The [https://www.solomotorcontrollers.com/ Solo Motor Controller] is a motor controller from Italy. [[Category:Stompy, Expand!]]" wikitext text/x-wiki The [https://www.solomotorcontrollers.com/ Solo Motor Controller] is a motor controller from Italy. [[Category:Stompy, Expand!]] 9aaf3ff505f9c0b8ec905a1c7197c456ba505b26 702 699 2024-04-30T00:34:19Z Admin 1 wikitext text/x-wiki The [https://www.solomotorcontrollers.com/ Solo Motor Controller] is a motor controller from Italy. [[Category:Stompy, Expand!]] [[Category:Actuators]] e23ca4c8a0ce1cf677f07dc090e904c1aea68561 705 702 2024-04-30T01:05:25Z Stompy 14 Bot expanded article wikitext text/x-wiki The '''Solo Motor Controller''' is a high-performance motor controller designed and manufactured in Italy. It is capable of handling various motor types, including DC, Brushless DC, Permanent Magnet Synchronous (PMSM), and AC induction motors. {{infobox actuator |name = Solo Motor Controller |manufacturer = Solo Motor Controllers |cost = |purchase_link = https://www.solomotorcontrollers.com/ |nominal_torque = |peak_torque = |weight = |dimensions = |gear_ratio = |voltage = 8 to 60VDC |cad_link = |interface = |gear_type = }} == Specifications == The Solo Motor Controller is distinguished by its compatibility with a broad spectrum of motors. It possesses advanced features and specifications. * **Voltage:** The Solo Motor Controller can operate with an input voltage from 8VDC to 60VDC. * **Continuous Operation:** With a continuous operation capacity of 110Arms, it can deliver up to a 120A DC current. * **Operation Modes:** The motor controller is designed to operate in a closed-loop method, enabling smooth and efficient motor control. However, detailed numerical specifications such as weight, dimensions, nominal and peak torque, gear ratio, and specifics of its interface are not readily available from mainstream sources. == Future Developments == The company promises continued advancement in their product line with the upcoming release of the "SOLO MEGA" in June 2023, a 6kW motor controller offering the same diverse motor compatibility but with even more enhanced features. == References == <references /> * [https://www.solomotorcontrollers.com/product/solo-uno/ Solo Motor Controllers - SOLO UNO] * [https://hackaday.io/project/170494-solo-a-motor-controller-for-all-motors Hackaday.io - SOLO, A Motor Controller for All Motors] * [https://www.solomotorcontrollers.com/resources/specs-datasheets/ Solo Motor Controllers - Specs & Datasheets] * [https://www.solomotorcontrollers.com/wp-content/uploads/materials/SOLO_MINI_User_Manual.pdf SOLO MINI User Manual] [[Category: Actuators]] 62c20ad56998d7099d0e7cb057519b59814c4d7d 707 705 2024-04-30T02:29:56Z 108.211.178.220 0 wikitext text/x-wiki The '''Solo Motor Controller''' is a high-performance motor controller designed and manufactured in Italy. It is capable of handling various motor types, including DC, Brushless DC, Permanent Magnet Synchronous (PMSM), and AC induction motors. {{infobox actuator |name = Solo Motor Controller |manufacturer = Solo Motor Controllers |cost = |purchase_link = https://www.solomotorcontrollers.com/ |nominal_torque = |peak_torque = |weight = |dimensions = |gear_ratio = |voltage = 8 to 60VDC |cad_link = |interface = |gear_type = }} == Specifications == The Solo Motor Controller is distinguished by its compatibility with a broad spectrum of motors. It possesses advanced features and specifications. * '''Voltage:''' The Solo Motor Controller can operate with an input voltage from 8VDC to 60VDC. * '''Continuous Operation:''' With a continuous operation capacity of 110Arms, it can deliver up to a 120A DC current. * '''Operation Modes:''' The motor controller is designed to operate in a closed-loop method, enabling smooth and efficient motor control. However, detailed numerical specifications such as weight, dimensions, nominal and peak torque, gear ratio, and specifics of its interface are not readily available from mainstream sources. == Future Developments == The company promises continued advancement in their product line with the upcoming release of the "SOLO MEGA" in June 2023, a 6kW motor controller offering the same diverse motor compatibility but with even more enhanced features. == References == <references /> * [https://www.solomotorcontrollers.com/product/solo-uno/ Solo Motor Controllers - SOLO UNO] * [https://hackaday.io/project/170494-solo-a-motor-controller-for-all-motors Hackaday.io - SOLO, A Motor Controller for All Motors] * [https://www.solomotorcontrollers.com/resources/specs-datasheets/ Solo Motor Controllers - Specs & Datasheets] * [https://www.solomotorcontrollers.com/wp-content/uploads/materials/SOLO_MINI_User_Manual.pdf SOLO MINI User Manual] [[Category: Actuators]] 93234a2a359ccaea7083234ed53d7eaa5f7e8993 0 170 701 2024-04-30T00:34:08Z Admin 1 Created page with "The [https://odriverobotics.com/ ODrive] is a precision motor controller. [[Category:Stompy, Expand!]] [[Category:Actuators]]" wikitext text/x-wiki The [https://odriverobotics.com/ ODrive] is a precision motor controller. [[Category:Stompy, Expand!]] [[Category:Actuators]] 3627f5ae4e6447e8496b2bee69d09a2866e4bf45 704 701 2024-04-30T01:04:57Z Stompy 14 Bot expanded article wikitext text/x-wiki Begin content expansion by calling the search function to procure further data about the ODrive Motor Controller. '''ODrive''' is a precision motor controller that is highly applicable in robotics domains, CNC machines, and other areas demanding high-performance motor control. It provides accurate control over hobby-grade and industrial electric motors. Here is detailed information procured through the Tavily Search API: ==Product Description== ODrive is primarily used for the control of electric motors with precision. The device is well-suited for accurate control of both hobby-grade and industrial motors. The controller has been specially designed for applications demanding high performance. The flexibility of the ODrive controller allows it to be used in a variety of applications including, but not limited to, robotics and CNC machines. ==Industrial Application== The ODrive motor controller finds its application in numerous sectors, particularly in robotics and CNC machines. This is mainly due to its flexibility, which facilitates control over a variety of motors ranging from hobby-grade to industrial ones. Furthermore, the precision and high performance offered by the controller make it a preferred choice for professionals. ==Technical Specifications== While the exact technical specifications of the ODrive motor controller can vary depending on the model, the controller is generally characterized by the following features: - High performance motor control: Ensures precision and accuracy. - Flexible interface: Facilitates easy integration with a variety of motors. - Compatible with various types of electric motors: Ensures wide application. {{infobox actuator | name = ODrive Motor Controller | manufacturer = ODrive Robotics | cost = To be confirmed | purchase_link = https://odriverobotics.com/ | nominal_torque = To be confirmed | peak_torque = To be confirmed | weight = To be confirmed | dimensions = To be confirmed | gear_ratio = N/A | voltage = To be confirmed | cad_link = To be confirmed | interface = To be confirmed | gear_type = N/A }} == References == <references /> [[Category:]] [[Category:Actuators]] 65d9479e430785c1482d650f5e9be8dc911243db 714 704 2024-04-30T03:53:08Z 108.211.178.220 0 wikitext text/x-wiki '''ODrive''' is a precision motor controller that is highly applicable in robotics domains, CNC machines, and other areas demanding high-performance motor control. It provides accurate control over hobby-grade and industrial electric motors. {{infobox actuator | name = ODrive Motor Controller | manufacturer = ODrive Robotics | cost = | purchase_link = https://odriverobotics.com/ | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = | gear_type = }} == Product Description == ODrive is primarily used for the control of electric motors with precision. The device is well-suited for accurate control of both hobby-grade and industrial motors. The controller has been specially designed for applications demanding high performance. The flexibility of the ODrive controller allows it to be used in a variety of applications including, but not limited to, robotics and CNC machines. == Industrial Application == The ODrive motor controller finds its application in numerous sectors, particularly in robotics and CNC machines. This is mainly due to its flexibility, which facilitates control over a variety of motors ranging from hobby-grade to industrial ones. Furthermore, the precision and high performance offered by the controller make it a preferred choice for professionals. == Technical Specifications == While the exact technical specifications of the ODrive motor controller can vary depending on the model, the controller is generally characterized by the following features: - High performance motor control: Ensures precision and accuracy. - Flexible interface: Facilitates easy integration with a variety of motors. - Compatible with various types of electric motors: Ensures wide application. == References == <references /> [[Category:]] [[Category:Actuators]] a98b91e32c1d8ccf85afb0608a91351fe7adfeca 715 714 2024-04-30T03:53:18Z 108.211.178.220 0 wikitext text/x-wiki '''ODrive''' is a precision motor controller that is highly applicable in robotics domains, CNC machines, and other areas demanding high-performance motor control. It provides accurate control over hobby-grade and industrial electric motors. {{infobox actuator | name = ODrive Motor Controller | manufacturer = ODrive Robotics | cost = | purchase_link = https://odriverobotics.com/ | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = | gear_type = }} == Product Description == ODrive is primarily used for the control of electric motors with precision. The device is well-suited for accurate control of both hobby-grade and industrial motors. The controller has been specially designed for applications demanding high performance. The flexibility of the ODrive controller allows it to be used in a variety of applications including, but not limited to, robotics and CNC machines. == Industrial Application == The ODrive motor controller finds its application in numerous sectors, particularly in robotics and CNC machines. This is mainly due to its flexibility, which facilitates control over a variety of motors ranging from hobby-grade to industrial ones. Furthermore, the precision and high performance offered by the controller make it a preferred choice for professionals. == Technical Specifications == While the exact technical specifications of the ODrive motor controller can vary depending on the model, the controller is generally characterized by the following features: * High performance motor control: Ensures precision and accuracy. * Flexible interface: Facilitates easy integration with a variety of motors. * Compatible with various types of electric motors: Ensures wide application. == References == <references /> [[Category:]] [[Category:Actuators]] 6a3f4dc8028d469c4f2a7e01062a1e6b61d09c93 0 41 706 651 2024-04-30T01:43:57Z 108.211.178.220 0 Add sections wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] PCB visualization software: Connecting Traces Exporting File for Manufacturing Exporting BoM == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] 8368d1253a0a0a872a73a5aebdd8005744105bcf 708 706 2024-04-30T02:42:08Z 108.211.178.220 0 add initial sections wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] PCB visualization software: Connecting Traces Exporting File for Manufacturing -Gerber Files -Drill Files -Map Files -BoM -CPL == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] b2fde67a40b08a23791a03bf0c5697d6ac34ae3f 709 708 2024-04-30T02:42:27Z 108.211.178.220 0 wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] PCB visualization software: Connecting Traces Exporting File for Manufacturing -Gerber Files -Drill Files -Map Files -BoM -CPL == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] df535f7d00946c3a90f42a7467d9fb766f65e2c0 710 709 2024-04-30T02:42:56Z 108.211.178.220 0 wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] PCB visualization software: Connecting Traces Exporting File for Manufacturing * Gerber Files * Drill Files * Map Files * BoM * CPL == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] 19e0cac56f251c4209539ea4a663672e7fa7013f 2 171 711 2024-04-30T02:43:40Z Matt 16 Initial Commit wikitext text/x-wiki Matt 0f9fe690f38da67968280971584cf9c16541f07b 712 711 2024-04-30T02:44:24Z Ben 2 wikitext text/x-wiki Matt? 666f830a57df374086bb652c333454e02fef3538 713 712 2024-04-30T02:44:58Z Ben 2 wikitext text/x-wiki Matt [[Category:Mazda Miata Enthusiasts]] 38616250f26fd5adbcd461cf412acf9250b6fa75 6 172 716 2024-04-30T04:49:50Z Matt 16 wikitext text/x-wiki KiCad Trace Tool a0a666c2b5b3bec72270c3521c031d20d5172346 0 16 717 502 2024-04-30T05:02:28Z 108.211.178.220 0 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Cluster 1 === === Cluster 2 === The cluster has 8 available nodes (each with 8 GPUs): <syntaxhighlight lang="text"> compute-permanent-node-68 compute-permanent-node-285 compute-permanent-node-493 compute-permanent-node-625 compute-permanent-node-626 compute-permanent-node-749 compute-permanent-node-801 compute-permanent-node-580 </syntaxhighlight> When you ssh-in, you log in to the bastion node pure-caribou-bastion from which you can log in to any other node where you can test your code. 55b46b4aa8353e0ee2f7659e997974e774b707b6 718 717 2024-04-30T05:05:24Z 108.211.178.220 0 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Cluster 1 === === Cluster 2 === The cluster has 8 available nodes (each with 8 GPUs): <syntaxhighlight lang="text"> compute-permanent-node-68 compute-permanent-node-285 compute-permanent-node-493 compute-permanent-node-625 compute-permanent-node-626 compute-permanent-node-749 compute-permanent-node-801 compute-permanent-node-580 </syntaxhighlight> When you ssh-in, you log in to the bastion node pure-caribou-bastion from which you can log in to any other node where you can test your code. [[Category:K-Scale]] 0e3d4eb093b5e27ee04f14052c610f64b7545786 14 173 719 2024-04-30T05:05:35Z 108.211.178.220 0 Created page with "Any documents related to K-Scale" wikitext text/x-wiki Any documents related to K-Scale 962f6bee8a7d9a538e9e0aa0a1038cee4a0a6a74 0 5 720 201 2024-04-30T05:05:56Z 108.211.178.220 0 wikitext text/x-wiki [[File:Logo.png|right|200px|thumb]] [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} [[Category:Companies]] [[Category:K-Scale]] 0003fc929362ae70b228c059e5709b560bdb5842 0 2 721 660 2024-04-30T05:06:03Z 108.211.178.220 0 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. [[Category:Robots]] [[Category:Open Source]] [[Category:K-Scale]] 50da2f5b2f67b5b4e961ff4826f8116d56978031 6 174 723 2024-04-30T05:22:39Z Matt 16 wikitext text/x-wiki JLCPCB BOM Format Requirement 8a4d49ab05368e0ba869f30ccfc60c4b696ddb7a 0 175 724 2024-04-30T05:23:36Z 108.211.178.220 0 Created page with "This document provides instructions on how to flash a Jetson Orin Nano or AGX. [[Category: Stompy, Expand!]]" wikitext text/x-wiki This document provides instructions on how to flash a Jetson Orin Nano or AGX. [[Category: Stompy, Expand!]] 82fb08432cc354db363d0df944f49991e483dcc3 10 176 726 2024-04-30T05:26:53Z 108.211.178.220 0 Created page with "{{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Introduced | value3 = {{{introduced_date|}}} |..." wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Introduced | value3 = {{{introduced_date|}}} | key4 = Standard | value4 = {{{standard|}}} | key5 = Network Type | value5 = {{{network_type|}}} | key6 = Topology | value6 = {{{topology|}}} | key7 = Speed | value7 = {{{speed|}}} | key8 = Distance | value8 = {{{distance|}}} | key9 = Website | value9 = {{#if: {{{website_link|}}} | [{{{website_link}}} Website] }} }} eefccfbb6d6a7cbc074071d9f2a7b2d3456bbf93 0 168 727 725 2024-04-30T05:27:08Z 108.211.178.220 0 wikitext text/x-wiki EtherCAT (Ethernet for Control Automation Technology) is a reliable, real-time Ethernet-based fieldbus system initially developed and implemented by Beckhoff Automation. Remarkably effective, this system uses a highly efficient communication method that allows data to be transferred from an industrial computer and communicated downstream to each piece of equipment in a daisy-chain configuration. The EtherCAT protocol is open and has gained wide acceptance due to its real-time capabilities and robust performance in industrial environments. {{infobox protocol | name = EtherCAT | organization = Beckhoff Automation | introduced_date = 2003 | standard = IEC 61158, IEC 61784-2 | network_type = Ethernet-based | topology = Daisy-chain, star, tree, or mixed arrangements | speed = Up to 100 Mbit/s | distance = Up to 100m (over copper cable), Up to 20km (over fiber) | website_link = https://www.ethercat.org/ }} The protocol was developed to offer high performance, real-time capabilities, and cost-effectiveness for systems that range from small to large scale. It finds a vast array of applications in machine controls, as well as in robotics and other fields that require high-speed, reliable communication between controllers and devices. This system is compatible with standard Ethernet and, therefore, does not require any specific hardware, which makes it a cost-effective solution. EtherCAT's ability to function on existing networks means it can work in tandem with other protocols and can share the same cables, switches, and the like, including industrial Ethernet infrastructure. == History == The EtherCAT technology was initially introduced by Beckhoff Automation in 2003. It was then smoothly transferred to the EtherCAT Technology Group (ETG) and has since been developed and maintained by this group. ETG is an open, non-profit organization that offers membership to manufacturers, developers, and users. The protocol has now established itself globally, becoming a high-performance Ethernet communication system in automation technology due to its real-time capabilities and the ability to connect a large number of devices to a single network. == Key Features of EtherCAT == EtherCAT offers several distinguishing features that contribute to its wide-spread use: * '''High Speed''': The system can process 1000 distributed I/O in 30 μs, or communicate with 100 servo axes in 100 μs. * '''Efficiency''': Data is not just passed along the communication line, but devices can directly read and write data as it flows by. * '''Robustness''': EtherCAT features error detection and flexible topology configurations to attain robustness against failure or data corruption. * '''Operability''': It allows for online configuration and diagnostics, making it a flexible and operable solution. * '''Cost-effectiveness''': Requires minimal hardware and can function over standard Ethernet, making its implementation cost-effective compared to other fieldbus systems. == Applications == The efficiency, robustness, and high speed of EtherCAT have made it a preferred choice in a host of applications: * '''Industrial Automation''': EtherCAT's high-speed, deterministic communication and robust error detection make it widely accepted in industrial automation, such as assembly lines, packing machinery, and material handling. * '''Robotics''': The perfect choice for complex servo control systems due to its capacity to handle fast control loops and synchronize large amounts of data. * '''Wind Energy''': EtherCAT is commonly used to control and monitor wind turbines due to its ability to accurately control multiple axes simultaneously. * '''Medical Technology''': In medical technology, EtherCAT is embraced for applications such as medical imaging, where it can handle the transfer of large amounts of data in real time. [[Category:Communication]] == References == <references /> 611ca1826f8ab54efafc70c088fc71d8f17f5a7d 728 727 2024-04-30T05:27:30Z 108.211.178.220 0 wikitext text/x-wiki EtherCAT (Ethernet for Control Automation Technology) is a reliable, real-time Ethernet-based fieldbus system initially developed and implemented by Beckhoff Automation. Remarkably effective, this system uses a highly efficient communication method that allows data to be transferred from an industrial computer and communicated downstream to each piece of equipment in a daisy-chain configuration. The EtherCAT protocol is open and has gained wide acceptance due to its real-time capabilities and robust performance in industrial environments. {{infobox protocol | name = EtherCAT | organization = Beckhoff Automation | introduced_date = 2003 | standard = IEC 61158, IEC 61784-2 | network_type = Ethernet-based | topology = Daisy-chain, star, tree, or mixed arrangements | speed = Up to 100 Mbit/s | distance = Up to 100m (over copper cable), Up to 20km (over fiber) | website_link = https://www.ethercat.org/ }} The protocol was developed to offer high performance, real-time capabilities, and cost-effectiveness for systems that range from small to large scale. It finds a vast array of applications in machine controls, as well as in robotics and other fields that require high-speed, reliable communication between controllers and devices. This system is compatible with standard Ethernet and, therefore, does not require any specific hardware, which makes it a cost-effective solution. EtherCAT's ability to function on existing networks means it can work in tandem with other protocols and can share the same cables, switches, and the like, including industrial Ethernet infrastructure. == History == The EtherCAT technology was initially introduced by Beckhoff Automation in 2003. It was then smoothly transferred to the EtherCAT Technology Group (ETG) and has since been developed and maintained by this group. ETG is an open, non-profit organization that offers membership to manufacturers, developers, and users. The protocol has now established itself globally, becoming a high-performance Ethernet communication system in automation technology due to its real-time capabilities and the ability to connect a large number of devices to a single network. == Key Features of EtherCAT == EtherCAT offers several distinguishing features that contribute to its wide-spread use: * '''High Speed''': The system can process 1000 distributed I/O in 30 μs, or communicate with 100 servo axes in 100 μs. * '''Efficiency''': Data is not just passed along the communication line, but devices can directly read and write data as it flows by. * '''Robustness''': EtherCAT features error detection and flexible topology configurations to attain robustness against failure or data corruption. * '''Operability''': It allows for online configuration and diagnostics, making it a flexible and operable solution. * '''Cost-effectiveness''': Requires minimal hardware and can function over standard Ethernet, making its implementation cost-effective compared to other fieldbus systems. == Applications == The efficiency, robustness, and high speed of EtherCAT have made it a preferred choice in a host of applications: * '''Industrial Automation''': EtherCAT's high-speed, deterministic communication and robust error detection make it widely accepted in industrial automation, such as assembly lines, packing machinery, and material handling. * '''Robotics''': The perfect choice for complex servo control systems due to its capacity to handle fast control loops and synchronize large amounts of data. * '''Wind Energy''': EtherCAT is commonly used to control and monitor wind turbines due to its ability to accurately control multiple axes simultaneously. * '''Medical Technology''': In medical technology, EtherCAT is embraced for applications such as medical imaging, where it can handle the transfer of large amounts of data in real time. [[Category:Communication]] 3e2fe1db9ed55dfee81e56797e7b8af3dd66c5ed 6 177 729 2024-04-30T05:31:48Z Matt 16 wikitext text/x-wiki settings for getting the CPL a7e600164c8cdc35eb38977780ac8d6d5ffa9959 6 178 730 2024-04-30T05:36:27Z Matt 16 wikitext text/x-wiki JLCPCB Example CPL 33a19765ef9ab790c0c3359a55c84b2e80e130b3 0 41 731 710 2024-04-30T05:41:54Z Matt 16 Add all information wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] === Importing into KiCad === After completing the atopile setup and building your atopile project, you will need to import the build into [https://www.kicad.org/ KiCad]. To import your design into KiCad, # Open <your-project>/elec/layout/default/<your-project-name>.kicad_pro with KiCad. # Delete all of the pcb board outlines you do not want to have (By default there are 3 available options to choose from) # Go to File->Import->Netlist... # In the Import Netlist pop-up, select your .net file to import, typically located at <your-project>/build/default.net # Click "Load and Test Netlist" # Click "Update PCB" # Click "Close" # Select anywhere on the screen where to place the components (you can move them later) atopile automatically connects the necessary components together, but you will still have to manually create your preferred layout and draw the connecting traces/routes (KiCad makes this process very simple) === Connecting Traces === After positioning your board components, you will have to connect them using the KiCad router tool, seen circled in red below: [[File:Trace Tool.png|center|thumb|KiCad Trace Tool]] To use this tool, simply select the router tool icon on the right-hand side of the KiCad program window and select a components pin to begin. The KiCad program will give you a visualization of which components you should trace towards. Connect all traces and verify no components have been left unconnected. === Exporting Files for Manufacturing === There are multiple files required to get a PCB manufactured. Each manufacturer may have different requirements. For this example, we will be using [https://jlcpcb.com/ JCLPCB's] PCB manufacturing services JCLPCB Requires: * Gerber Files * Drill Files * Map Files * BoM * CPL ==== Exporting Gerber, Drill, and Map Files ==== For a detailed instruction follow JLCPCB's KiCad export instructions [https://jlcpcb.com/help/article/362-how-to-generate-gerber-and-drill-files-in-kicad-7 here] ==== Exporting BoM (Bill of Materials) ==== atopile will automatically make the BoM for you, although you may need to reformat the header & information to meet the requirements of your manufacturer. The JLCPCP format can be seen below: [[File:JLCPCB BOM Format.png|center|thumb|JLCPCB BOM Requirement]] atopile's BoM file can be found in the build directory, typically called "default.csv" (<your-project>/build/default.net) ==== Exporting CPL (Component Placement List) ==== KiCad allows for quick and easy CPL exporting, although you will have to reformat your information to fit your manufacturers requirements To export a CPL from your KiCad project: # Go to File -> Fabrication Outputs -> Component Placement # Select proper output directory # Use CSV, Millimeters, and Separate files for front, back settings: [[File:Pos settings.png|center|thumb|Settings for CPL Generation in KiCad]] # Click "Generate Position File" # Fix output file to match your manufacturers requirements (JLCPCB example provided) [[File:JLCPCB CPL.png|center|thumb|JLCPCB CPL Example Format]] == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] a7a7d8d603619ba8cd9d4647d53351a91a6e2189 732 731 2024-04-30T05:44:06Z Matt 16 space wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] === Importing into KiCad === After completing the atopile setup and building your atopile project, you will need to import the build into [https://www.kicad.org/ KiCad]. To import your design into KiCad, # Open <your-project>/elec/layout/default/<your-project-name>.kicad_pro with KiCad. # Delete all of the pcb board outlines you do not want to have (By default there are 3 available options to choose from) # Go to File->Import->Netlist... # In the Import Netlist pop-up, select your .net file to import, typically located at <your-project>/build/default.net # Click "Load and Test Netlist" # Click "Update PCB" # Click "Close" # Select anywhere on the screen where to place the components (you can move them later) atopile automatically connects the necessary components together, but you will still have to manually create your preferred layout and draw the connecting traces/routes (KiCad makes this process very simple) === Connecting Traces === After positioning your board components, you will have to connect them using the KiCad router tool, seen circled in red below: [[File:Trace Tool.png|center|thumb|KiCad Trace Tool]] To use this tool, simply select the router tool icon on the right-hand side of the KiCad program window and select a components pin to begin. The KiCad program will give you a visualization of which components you should trace towards. Connect all traces and verify no components have been left unconnected. === Exporting Files for Manufacturing === There are multiple files required to get a PCB manufactured. Each manufacturer may have different requirements. For this example, we will be using [https://jlcpcb.com/ JCLPCB's] PCB manufacturing services JCLPCB Requires: * Gerber Files * Drill Files * Map Files * BoM * CPL ==== Exporting Gerber, Drill, and Map Files ==== For a detailed instruction follow JLCPCB's KiCad export instructions [https://jlcpcb.com/help/article/362-how-to-generate-gerber-and-drill-files-in-kicad-7 here] ==== Exporting BoM (Bill of Materials) ==== atopile will automatically make the BoM for you, although you may need to reformat the header & information to meet the requirements of your manufacturer. The JLCPCP format can be seen below: [[File:JLCPCB BOM Format.png|center|thumb|JLCPCB BOM Requirement]] atopile's BoM file can be found in the build directory, typically called "default.csv" (<your-project>/build/default.net) ==== Exporting CPL (Component Placement List) ==== KiCad allows for quick and easy CPL exporting, although you will have to reformat your information to fit your manufacturers requirements To export a CPL from your KiCad project: # Go to File -> Fabrication Outputs -> Component Placement # Select proper output directory # Use CSV, Millimeters, and Separate files for front, back settings: [[File:Pos settings.png|center|thumb|Settings for CPL Generation in KiCad]] # Click "Generate Position File" # Fix output file to match your manufacturers requirements (JLCPCB example provided) [[File:JLCPCB CPL.png|center|thumb|JLCPCB CPL Example Format]] == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] 11af264c9970a896215de47fc31053cb348980aa 733 732 2024-04-30T05:44:46Z 108.211.178.220 0 wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] === Importing into KiCad === After completing the atopile setup and building your atopile project, you will need to import the build into [https://www.kicad.org/ KiCad]. To import your design into KiCad, # Open <your-project>/elec/layout/default/<your-project-name>.kicad_pro with KiCad. # Delete all of the pcb board outlines you do not want to have (By default there are 3 available options to choose from) # Go to File->Import->Netlist... # In the Import Netlist pop-up, select your .net file to import, typically located at <your-project>/build/default.net # Click "Load and Test Netlist" # Click "Update PCB" # Click "Close" # Select anywhere on the screen where to place the components (you can move them later) atopile automatically connects the necessary components together, but you will still have to manually create your preferred layout and draw the connecting traces/routes (KiCad makes this process very simple) === Connecting Traces === After positioning your board components, you will have to connect them using the KiCad router tool, seen circled in red below: [[File:Trace Tool.png|center|thumb|KiCad Trace Tool]] To use this tool, simply select the router tool icon on the right-hand side of the KiCad program window and select a components pin to begin. The KiCad program will give you a visualization of which components you should trace towards. Connect all traces and verify no components have been left unconnected. === Exporting Files for Manufacturing === There are multiple files required to get a PCB manufactured. Each manufacturer may have different requirements. For this example, we will be using [https://jlcpcb.com/ JCLPCB's] PCB manufacturing services JCLPCB Requires: * Gerber Files * Drill Files * Map Files * BoM * CPL ==== Exporting Gerber, Drill, and Map Files ==== For a detailed instruction follow JLCPCB's KiCad export instructions [https://jlcpcb.com/help/article/362-how-to-generate-gerber-and-drill-files-in-kicad-7 here] ==== Exporting BoM (Bill of Materials) ==== atopile will automatically make the BoM for you, although you may need to reformat the header & information to meet the requirements of your manufacturer. The JLCPCP format can be seen below: [[File:JLCPCB BOM Format.png|center|500px|thumb|JLCPCB BOM Requirement]] atopile's BoM file can be found in the build directory, typically called "default.csv" (<your-project>/build/default.net) ==== Exporting CPL (Component Placement List) ==== KiCad allows for quick and easy CPL exporting, although you will have to reformat your information to fit your manufacturers requirements To export a CPL from your KiCad project: # Go to File -> Fabrication Outputs -> Component Placement # Select proper output directory # Use CSV, Millimeters, and Separate files for front, back settings: [[File:Pos settings.png|center|500px|thumb|Settings for CPL Generation in KiCad]] # Click "Generate Position File" # Fix output file to match your manufacturers requirements (JLCPCB example provided) [[File:JLCPCB CPL.png|center|500px|thumb|JLCPCB CPL Example Format]] == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] f5b357e6ddbfe774b9f34a46d04c172ba472db00 734 733 2024-04-30T05:45:08Z Matt 16 Gram wikitext text/x-wiki Walk-through and notes regarding how to design and ship a PCB. == Designing with atopile == [[atopile]] enables code-defined pcb design. Follow atopile's [https://atopile.io/getting-started/ getting-started] guide to set up your project. An example atopile PCB project is provided by the K-Scale Labs team [https://github.com/kscalelabs/atopile-pcb-example here] === Importing into KiCad === After completing the atopile setup and building your atopile project, you will need to import the build into [https://www.kicad.org/ KiCad]. To import your design into KiCad, # Open <your-project>/elec/layout/default/<your-project-name>.kicad_pro with KiCad. # Delete all of the pcb board outlines you do not want to have (By default there are 3 available options to choose from) # Go to File->Import->Netlist... # In the Import Netlist pop-up, select your .net file to import, typically located at <your-project>/build/default.net # Click "Load and Test Netlist" # Click "Update PCB" # Click "Close" # Select anywhere on the screen where to place the components (you can move them later) atopile automatically connects the necessary components together, but you will still have to manually create your preferred layout and draw the connecting traces/routes (KiCad makes this process very simple) === Connecting Traces === After positioning your board components, you will have to connect them using the KiCad router tool, seen circled in red below: [[File:Trace Tool.png|center|thumb|KiCad Trace Tool]] To use this tool, simply select the router tool icon on the right-hand side of the KiCad program window and select a components pin to begin. The KiCad program will give you a visualization of which components you should trace towards. Connect all traces and verify no components have been left unconnected. === Exporting Files for Manufacturing === There are multiple files required to get a PCB manufactured. Each manufacturer may have different requirements. For this example, we will be using [https://jlcpcb.com/ JCLPCB's] PCB manufacturing services. JCLPCB Requires: * Gerber Files * Drill Files * Map Files * BoM * CPL ==== Exporting Gerber, Drill, and Map Files ==== For a detailed instruction, follow JLCPCB's KiCad export instructions [https://jlcpcb.com/help/article/362-how-to-generate-gerber-and-drill-files-in-kicad-7 here] ==== Exporting BoM (Bill of Materials) ==== atopile will automatically make the BoM for you, although you may need to reformat the header & information to meet the requirements of your manufacturer. The JLCPCP format can be seen below: [[File:JLCPCB BOM Format.png|center|500px|thumb|JLCPCB BOM Requirement]] atopile's BoM file can be found in the build directory, typically called "default.csv" (<your-project>/build/default.net) ==== Exporting CPL (Component Placement List) ==== KiCad allows for quick and easy CPL exporting, although you will have to reformat your information to fit your manufacturers requirements To export a CPL from your KiCad project: # Go to File -> Fabrication Outputs -> Component Placement # Select proper output directory # Use CSV, Millimeters, and Separate files for front, back settings: [[File:Pos settings.png|center|500px|thumb|Settings for CPL Generation in KiCad]] # Click "Generate Position File" # Fix output file to match your manufacturers requirements (JLCPCB example provided) [[File:JLCPCB CPL.png|center|500px|thumb|JLCPCB CPL Example Format]] == Ordering a PCB == Trusted low production PCB manufacturing companies: * PCBWay * JLCPCB * SeeedStudio Further PCB manufacturers and price comparisons for your specific project can be found [https://pcbshopper.com/ here] == Related Articles == * [[atopile]] [[Category: Hardware]] [[Category: Guides]] [[Category: Electronics]] 7488963f2cfc5d65254a65c14ec8d215bc4e5c48 2 171 735 713 2024-04-30T05:51:41Z Matt 16 Add info wikitext text/x-wiki Matt is the co-founder and hardware lead of [[K-Scale Labs]]. {{infobox person | name = Matt Freed | organization = [[K-Scale Labs]] | title = Head of Hardware }} [[Category: K-Scale Employees]] e6ebcfc5afd9e20344d161fcfad4289da93b380a 0 175 736 724 2024-04-30T06:04:06Z Stompy 14 Bot expanded article wikitext text/x-wiki == Overview == Flashing a Jetson Orin Nano or AGX is a process that involves updating the software or firmware installed on these NVIDIA Jetson models. It's a common technique used by developers to either install a new system or restore an existing one. Please follow the below steps to successfully flash your device. {{infobox company | name = NVIDIA | country = United States | website_link = https://www.nvidia.com/ | robots = Jetson Orin Nano, Jetson AGX Xavier }} == Prerequisites == Before beginning the flashing process, ensure you have the following: * NVIDIA's flashing software, such as the NVIDIA SDK Manager * A compatible host machine (either Windows, Linux, or Mac) * USB cable to connect the Jetson device to the host machine * Jetson Orin Nano or AGX device == Flashing Process == The following are the general steps to follow when flashing a Jetson Orin Nano or AGX: # Download and install the appropriate NVIDIA SDK Manager on your host machine. # Connect the Jetson device to your host machine using the USB cable. # Open the NVIDIA SDK Manager and select your Jetson device (either Orin Nano or AGX). # Follow the on-screen instructions to complete the flashing process. Once the flashing process is complete, your Jetson device should now be updated with the latest firmware or software. == Troubleshooting == If the flashing process fails or encounters any issues, refer to NVIDIA's official documentation or contact their support for assistance. == References == <references /> [[Category:Robotics]] [[Category:Instruction]] Functions used: '''functions.search_web''' Query: "NVIDIA Jetson Orin Nano or AGX flashing instructions" ```json { "query": "NVIDIA Jetson Orin Nano or AGX flashing instructions" } ``` 42dbd469b9d0697a50c07aaa8dbe653cb086affd 14 60 737 427 2024-04-30T12:01:41Z 136.62.52.52 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com * https://github.com/wuphilipp/gello_software * https://www.ais.uni-bonn.de/papers/Humanoids_2023_Schwarz_Lenz_NimbRo_Avatar.pdf * https://robot.neu.edu/project/avatar/ == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC, WebRTC, ZMQ. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com * https://github.com/ToruOwO/hato/tree/main?tab=readme-ov-file#collecting-demonstration-data * https://github.com/rail-berkeley/oculus_reader * https://github.com/OpenTeleVision/TeleVision/tree/main == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] 8a99f62058b0891abdb065970ba6a09facf9c0d6 738 737 2024-04-30T12:03:24Z 136.62.52.52 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com * https://github.com/wuphilipp/gello_software * https://www.ais.uni-bonn.de/papers/Humanoids_2023_Schwarz_Lenz_NimbRo_Avatar.pdf * https://www.ais.uni-bonn.de/papers/SORO_2023_Lenz_Schwarz_NimbRo_Avatar.pdf * https://robot.neu.edu/project/avatar/ == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC, WebRTC, ZMQ. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com * https://github.com/ToruOwO/hato/tree/main?tab=readme-ov-file#collecting-demonstration-data * https://github.com/rail-berkeley/oculus_reader * https://github.com/OpenTeleVision/TeleVision/tree/main == Controller Teleop == In controlelr teleop, the user uses joysticks, keyborads, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] d1e7ed934a875a9e0657fa03cb506c36d03b111c 739 738 2024-04-30T16:52:28Z 108.211.178.220 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com * https://github.com/wuphilipp/gello_software * https://www.ais.uni-bonn.de/papers/Humanoids_2023_Schwarz_Lenz_NimbRo_Avatar.pdf * https://www.ais.uni-bonn.de/papers/SORO_2023_Lenz_Schwarz_NimbRo_Avatar.pdf * https://robot.neu.edu/project/avatar/ == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC, WebRTC, ZMQ. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com * https://github.com/ToruOwO/hato/tree/main?tab=readme-ov-file#collecting-demonstration-data * https://github.com/rail-berkeley/oculus_reader * https://github.com/OpenTeleVision/TeleVision/tree/main == Controller Teleop == In controller teleop, the user uses joysticks, keyboards, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 [[Category: Teleop]] ccb19761d033461f7ce2cb5ccf84e5007955bbcc 740 739 2024-04-30T16:54:10Z 108.211.178.220 0 wikitext text/x-wiki = Teleop = Teleoperation is the art of controlling a robot from a distance (prefix "tele-" comes from the Ancient Greek word tēle, which means "far off, at a distance, far away, far from"). Robots specifically designed to be teleoperated by a human are known as a Proxy. == Whole Body Teleop == In whole-body teleop, the user controls a hardware replica of the robot directly. The real robot directly mimcis the pose of the replica robot. * https://mobile-aloha.github.io/ * https://www.youtube.com/watch?v=PFw5hwNVhbA * https://x.com/haoshu_fang/status/1707434624413306955 * https://x.com/aditya_oberai/status/1762637503495033171 * https://what-is-proxy.com * https://github.com/wuphilipp/gello_software * https://www.ais.uni-bonn.de/papers/Humanoids_2023_Schwarz_Lenz_NimbRo_Avatar.pdf * https://www.ais.uni-bonn.de/papers/SORO_2023_Lenz_Schwarz_NimbRo_Avatar.pdf * https://robot.neu.edu/project/avatar/ == VR Teleop == In VR teleop the user controls a simulated version of the robot. The robot and VR headset are usually on the same LAN and communication is done via gRPC, WebRTC, ZMQ. * https://github.com/Improbable-AI/VisionProTeleop * https://github.com/fazildgr8/VR_communication_mujoco200 * https://github.com/pollen-robotics/reachy2021-unity-package * https://freetale.medium.com/unity-grpc-in-2023-98b739cb115 * https://x.com/AndreTI/status/1780665435999924343 * https://holo-dex.github.io/ * https://what-is-proxy.com * https://github.com/ToruOwO/hato/tree/main?tab=readme-ov-file#collecting-demonstration-data * https://github.com/rail-berkeley/oculus_reader * https://github.com/OpenTeleVision/TeleVision/tree/main == Controller Teleop == In controller teleop, the user uses joysticks, keyboards, and other controllers to control the robot. The buttons map to hardcoded behaviors on the robot. * https://github.com/ros-teleop * https://x.com/ShivinDass/status/1606156894271197184 * https://what-is-proxy.com == Latency == * https://ennerf.github.io/2016/09/20/A-Practical-Look-at-Latency-in-Robotics-The-Importance-of-Metrics-and-Operating-Systems.html * https://link.springer.com/article/10.1007/s10846-022-01749-3 * https://twitter.com/watneyrobotics/status/1769058250731999591 26e6a7febfc97297d71cb02510cbf438de80035f 2 141 743 488 2024-04-30T17:47:16Z Ben 2 Protected "[[User:Ben]]" ([Edit=Allow only administrators] (indefinite) [Move=Allow only administrators] (indefinite)) wikitext text/x-wiki [[File:Ben.jpg|right|200px|thumb]] [https://ben.bolte.cc/ Ben] is the founder and CEO of [[K-Scale Labs]]. {{infobox person | name = Ben Bolte | organization = [[K-Scale Labs]] | title = CEO | website_link = https://ben.bolte.cc/ }} [[Category: K-Scale Employees]] f971d4004fa84319343a24548245ebcf11a88d16 0 180 747 2024-04-30T18:47:31Z Ben 2 Created page with "Project documentation for the open-source teleoperated arm from [[K-Scale Labs]]. [[Category:K-Scale]]" wikitext text/x-wiki Project documentation for the open-source teleoperated arm from [[K-Scale Labs]]. [[Category:K-Scale]] 73b2044943cc63dd975102d143161221135a9154 0 181 748 2024-04-30T19:07:12Z Ben 2 Created page with "The '''K-Scale Onshape Library'''<ref>https://github.com/kscalelabs/onshape</ref> is a tool developed by [[K-Scale Labs]] to convert Onshape files into simulation artifacts...." wikitext text/x-wiki The '''K-Scale Onshape Library'''<ref>https://github.com/kscalelabs/onshape</ref> is a tool developed by [[K-Scale Labs]] to convert Onshape files into simulation artifacts. === Sample Script === Below is a sample script for converting a [[Stompy]] Onshape model to a URDF. <syntaxhighlight lang="bash> #!/bin/zsh # ./sim/scripts/download_urdf.sh # URL of the latest model. url=https://cad.onshape.com/documents/71f793a23ab7562fb9dec82d/w/6160a4f44eb6113d3fa116cd/e/1a95e260677a2d2d5a3b1eb3 # Output directory. output_dir=${MODEL_DIR}/robots/stompy kol urdf ${url} \ --max-ang-velocity 31.4 \ --suffix-to-joint-effort \ dof_x4_h=1.5 \ dof_x4=1.5 \ dof_x6=3 \ dof_x8=6 \ dof_x10=12 \ knee_revolute=13.9 \ ankle_revolute=6 \ --output-dir ${output_dir} \ --disable-mimics \ --mesh-ext obj </syntaxhighlight> 61f7782f7c1a403609f78c657bb6c8bee8b1cbc7 749 748 2024-04-30T19:07:28Z Ben 2 wikitext text/x-wiki The '''K-Scale Onshape Library'''<ref>https://github.com/kscalelabs/onshape</ref> is a tool developed by [[K-Scale Labs]] to convert Onshape files into simulation artifacts. === Sample Script === Below is a sample script for converting a [[Stompy]] Onshape model to a URDF<ref>https://github.com/kscalelabs/sim/blob/master/sim/scripts/download_urdf.sh</ref>. <syntaxhighlight lang="bash> #!/bin/zsh # ./sim/scripts/download_urdf.sh # URL of the latest model. url=https://cad.onshape.com/documents/71f793a23ab7562fb9dec82d/w/6160a4f44eb6113d3fa116cd/e/1a95e260677a2d2d5a3b1eb3 # Output directory. output_dir=${MODEL_DIR}/robots/stompy kol urdf ${url} \ --max-ang-velocity 31.4 \ --suffix-to-joint-effort \ dof_x4_h=1.5 \ dof_x4=1.5 \ dof_x6=3 \ dof_x8=6 \ dof_x10=12 \ knee_revolute=13.9 \ ankle_revolute=6 \ --output-dir ${output_dir} \ --disable-mimics \ --mesh-ext obj </syntaxhighlight> a43b10e0f29da1f3ff922079a7ca806cbffab3df 750 749 2024-04-30T19:07:33Z Ben 2 wikitext text/x-wiki The '''K-Scale Onshape Library'''<ref>https://github.com/kscalelabs/onshape</ref> is a tool developed by [[K-Scale Labs]] to convert Onshape files into simulation artifacts. === Sample Script === Below is a sample script for converting a [[Stompy]] Onshape model to a URDF<ref>https://github.com/kscalelabs/sim/blob/master/sim/scripts/download_urdf.sh</ref>. <syntaxhighlight lang="bash> #!/bin/zsh # URL of the latest model. url=https://cad.onshape.com/documents/71f793a23ab7562fb9dec82d/w/6160a4f44eb6113d3fa116cd/e/1a95e260677a2d2d5a3b1eb3 # Output directory. output_dir=${MODEL_DIR}/robots/stompy kol urdf ${url} \ --max-ang-velocity 31.4 \ --suffix-to-joint-effort \ dof_x4_h=1.5 \ dof_x4=1.5 \ dof_x6=3 \ dof_x8=6 \ dof_x10=12 \ knee_revolute=13.9 \ ankle_revolute=6 \ --output-dir ${output_dir} \ --disable-mimics \ --mesh-ext obj </syntaxhighlight> 7b5394eed1a93cbf30bae8ef4553ac6bd66df66f 0 155 751 686 2024-04-30T22:45:53Z 108.211.178.220 0 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> ==== MCP2515 Driver ==== == References == <references /> [[Category:Communication]] 81460e77f1384b55c059a1ce2b494022760b030e 768 751 2024-05-01T01:27:26Z Budzianowski 19 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. === MCP2515 === The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> === Applications === The [https://python-can.readthedocs.io/en/stable/ python-can] library provides Controller Area Network support for Python, providing common abstractions to different hardware devices, and a suite of utilities for sending and receiving messages on a CAN bus. There is an example of an installation on [[Jetson_Orin]] here. ==== Pi4 ==== Raspberry Pi offers an easy to deploy 2-Channel Isolated CAN Bus Expansion HAT which allows to quickly integrate it to the peripheral devices. See the [https://www.waveshare.com/wiki/2-CH_CAN_HAT tutorial] for more information === Arduino === Arduino has a good support of the MCPs with many implementations of the [https://github.com/Seeed-Studio/Seeed_Arduino_CAN drivers] === MCP2515 Driver === By default the CAN bus node is supposed to acknowledge every message on the bus weather or not that node is interested in the message. However, the interference on the network can drop some bits during the communication. In the standard mode, the node would not only continuously try to re-send the unacknowledged messages, but also after a short period it would start sending error frames and then eventually go to bus-off mode and stop. This causes sever issues when the CAN network works with multiple motors. The controller has a [http://ww1.microchip.com/downloads/en/DeviceDoc/MCP2515-Stand-Alone-CAN-Controller-with-SPI-20001801J.pdf one-shot] setup that requires changes in the driver. == References == <references /> [[Category:Communication]] f4701a3d65b6ae701b5e1a2f6d155536c2b6abb9 769 768 2024-05-01T01:28:05Z Budzianowski 19 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. == MCP2515 == The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == Applications == The [https://python-can.readthedocs.io/en/stable/ python-can] library provides Controller Area Network support for Python, providing common abstractions to different hardware devices, and a suite of utilities for sending and receiving messages on a CAN bus. There is an example of an installation on [[Jetson_Orin]] here. === Pi4 === Raspberry Pi offers an easy to deploy 2-Channel Isolated CAN Bus Expansion HAT which allows to quickly integrate it to the peripheral devices. See the [https://www.waveshare.com/wiki/2-CH_CAN_HAT tutorial] for more information === Arduino === Arduino has a good support of the MCPs with many implementations of the [https://github.com/Seeed-Studio/Seeed_Arduino_CAN drivers] === MCP2515 Driver === By default the CAN bus node is supposed to acknowledge every message on the bus weather or not that node is interested in the message. However, the interference on the network can drop some bits during the communication. In the standard mode, the node would not only continuously try to re-send the unacknowledged messages, but also after a short period it would start sending error frames and then eventually go to bus-off mode and stop. This causes sever issues when the CAN network works with multiple motors. The controller has a [http://ww1.microchip.com/downloads/en/DeviceDoc/MCP2515-Stand-Alone-CAN-Controller-with-SPI-20001801J.pdf one-shot] setup that requires changes in the driver. == References == <references /> [[Category:Communication]] ba36d170b88c7639fdc62ee3e3294b9d7389da4b 770 769 2024-05-01T01:28:28Z Budzianowski 19 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. == MCP2515 == The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == Applications == The [https://python-can.readthedocs.io/en/stable/ python-can] library provides Controller Area Network support for Python, providing common abstractions to different hardware devices, and a suite of utilities for sending and receiving messages on a CAN bus. There is an example of an installation on [[Jetson_Orin]] here. === Pi4 === Raspberry Pi offers an easy to deploy 2-Channel Isolated CAN Bus Expansion HAT which allows to quickly integrate it to the peripheral devices. See the [https://www.waveshare.com/wiki/2-CH_CAN_HAT tutorial] for more information === Arduino === Arduino has a good support of the MCPs with many implementations of the [https://github.com/Seeed-Studio/Seeed_Arduino_CAN drivers] === MCP2515 Driver === By default the CAN bus node is supposed to acknowledge every message on the bus weather or not that node is interested in the message. However, the interference on the network can drop some bits during the communication. In the standard mode, the node would not only continuously try to re-send the unacknowledged messages, but also after a short period it would start sending error frames and then eventually go to bus-off mode and stop. This causes sever issues when the CAN network works with multiple motors. The controller has a [http://ww1.microchip.com/downloads/en/DeviceDoc/MCP2515-Stand-Alone-CAN-Controller-with-SPI-20001801J.pdf one-shot] setup that requires changes in the driver. == References == <references /> [[Category:Communication]] b54b57ad3e3f520c864439646564f9579fb1435e 771 770 2024-05-01T01:28:52Z Budzianowski 19 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. == MCP2515 == The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515</ref> Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet]. The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == Applications == The [https://python-can.readthedocs.io/en/stable/ python-can] library provides Controller Area Network support for Python, providing common abstractions to different hardware devices, and a suite of utilities for sending and receiving messages on a CAN bus. There is an example of an installation on [[Jetson_Orin]] here. === Pi4 === Raspberry Pi offers an easy to deploy 2-Channel Isolated CAN Bus Expansion HAT which allows to quickly integrate it to the peripheral devices. See the [https://www.waveshare.com/wiki/2-CH_CAN_HAT tutorial] for more information === Arduino === Arduino has a good support of the MCPs with many implementations of the [https://github.com/Seeed-Studio/Seeed_Arduino_CAN drivers] === MCP2515 Driver === By default the CAN bus node is supposed to acknowledge every message on the bus weather or not that node is interested in the message. However, the interference on the network can drop some bits during the communication. In the standard mode, the node would not only continuously try to re-send the unacknowledged messages, but also after a short period it would start sending error frames and then eventually go to bus-off mode and stop. This causes sever issues when the CAN network works with multiple motors. The controller has a [http://ww1.microchip.com/downloads/en/DeviceDoc/MCP2515-Stand-Alone-CAN-Controller-with-SPI-20001801J.pdf one-shot] setup that requires changes in the driver. == References == <references /> [[Category:Communication]] e69e7301f068d5801538a432910f0b3591285686 772 771 2024-05-01T01:29:06Z Budzianowski 19 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. == MCP2515 == The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == Applications == The [https://python-can.readthedocs.io/en/stable/ python-can] library provides Controller Area Network support for Python, providing common abstractions to different hardware devices, and a suite of utilities for sending and receiving messages on a CAN bus. There is an example of an installation on [[Jetson_Orin]] here. === Pi4 === Raspberry Pi offers an easy to deploy 2-Channel Isolated CAN Bus Expansion HAT which allows to quickly integrate it to the peripheral devices. See the [https://www.waveshare.com/wiki/2-CH_CAN_HAT tutorial] for more information === Arduino === Arduino has a good support of the MCPs with many implementations of the [https://github.com/Seeed-Studio/Seeed_Arduino_CAN drivers] === MCP2515 Driver === By default the CAN bus node is supposed to acknowledge every message on the bus weather or not that node is interested in the message. However, the interference on the network can drop some bits during the communication. In the standard mode, the node would not only continuously try to re-send the unacknowledged messages, but also after a short period it would start sending error frames and then eventually go to bus-off mode and stop. This causes sever issues when the CAN network works with multiple motors. The controller has a [http://ww1.microchip.com/downloads/en/DeviceDoc/MCP2515-Stand-Alone-CAN-Controller-with-SPI-20001801J.pdf one-shot] setup that requires changes in the driver. == References == <references /> [[Category:Communication]] b54b57ad3e3f520c864439646564f9579fb1435e 0 182 752 2024-05-01T00:06:01Z Budzianowski 19 Created page with "A guide to simulate Stompy walking. [[Category: Guides]] [[Category: Software]]" wikitext text/x-wiki A guide to simulate Stompy walking. [[Category: Guides]] [[Category: Software]] 71b1225810ae841c4e9ed317963cbc926cec27bf 755 752 2024-05-01T00:06:43Z Budzianowski 19 Budzianowski moved page [[Simulation guide]] to [[Walking Stompy guide]] wikitext text/x-wiki A guide to simulate Stompy walking. [[Category: Guides]] [[Category: Software]] 71b1225810ae841c4e9ed317963cbc926cec27bf 10 183 753 2024-05-01T00:06:28Z Ben 2 Created page with "{{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Full Name | value2 = {{{full_name}}} | key3 = Arxiv | value3 = {{#if: {{{arxiv_link|}}} | [{{{arxiv_..." wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Full Name | value2 = {{{full_name}}} | key3 = Arxiv | value3 = {{#if: {{{arxiv_link|}}} | [{{{arxiv_link}}} Link] }} | key4 = Project Page | value4 = {{#if: {{{project_link|}}} | [{{{project_link}}} Website] }} | key5 = Twitter | value5 = {{#if: {{{twitter_link|}}} | [{{{twitter_link}}} Twitter] }} | key6 = Publication Date | value6 = {{{date|}}} | key7 = Authors | value7 = {{{authors|}}} }} d535150cf5288a390695971df41697349e1774bc 754 753 2024-05-01T00:06:32Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Full Name | value2 = {{{full_name|}}} | key3 = Arxiv | value3 = {{#if: {{{arxiv_link|}}} | [{{{arxiv_link}}} Link] }} | key4 = Project Page | value4 = {{#if: {{{project_link|}}} | [{{{project_link}}} Website] }} | key5 = Twitter | value5 = {{#if: {{{twitter_link|}}} | [{{{twitter_link}}} Twitter] }} | key6 = Publication Date | value6 = {{{date|}}} | key7 = Authors | value7 = {{{authors|}}} }} 61ecf4fcda3f2f391640a2d7ca47a3ac0742fec8 0 184 756 2024-05-01T00:06:43Z Budzianowski 19 Budzianowski moved page [[Simulation guide]] to [[Walking Stompy guide]] wikitext text/x-wiki #REDIRECT [[Walking Stompy guide]] 843539494d81f842a9d35875aa1afdf1c8aa05e8 0 66 757 289 2024-05-01T00:07:39Z Budzianowski 19 wikitext text/x-wiki =Applications List= A non-comprehensive list of training frameworks is listed below. ===[https://github.com/leggedrobotics/legged_gym Legged Gym]=== Isaac Gym Environments for Legged Robots. ===[https://github.com/chengxuxin/extreme-parkour Extreme Parkour]=== Extreme Parkour with AMP Legged Robots. ===[https://github.com/Alescontrela/AMP_for_hardware AMP for hardware]=== Adversarial Motion Priors Make Good Substitutes for Complex Reward Functions ===[https://github.com/ZhengyiLuo/PHC PHC]=== Official Implementation of the ICCV 2023 paper: Perpetual Humanoid Control for Real-time Simulated Avatars. ===[https://github.com/roboterax/humanoid-gym Humanoid Gym]=== Training setup for walking with [[Xbot-L]]. ===[https://github.com/kscalelabs/sim KScale Sim]=== Training setup for getting up and walking with [[Stompy]]. [[Category: Software]] 2a756c2265c456337d18f09f69c22c02f74ed71b 0 18 758 342 2024-05-01T00:08:04Z Budzianowski 19 wikitext text/x-wiki Isaac Sim is a simulator from NVIDIA connect with the Omniverse platform. The core physics engine underlying Isaac Sim is PhysX === Doing Simple Operations === '''Start Isaac Sim''' * Open Omniverse Launcher * Navigate to the Library * Under “Apps” click “Isaac Sim” * Click “Launch” ** There are multiple options for launching. Choose the normal one to show the GUI or headless if streaming. * Choose <code>File > Open...</code> and select the <code>.usd</code> model corresponding to the robot you want to simulate. '''Connecting streaming client''' * Start Isaac Sim in Headless (Native) mode * Open Omniverse Streaming Client * Connect to the server [[Category: Software]] [[Category: Simulators]] ee307ceaaa4f5a2d87fa7daecc75b7a1a453db5c 0 34 759 628 2024-05-01T00:08:25Z Budzianowski 19 wikitext text/x-wiki == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] [[Category: Software]] 81ff653e50db6b0b9d2db8edbb11a9b7c94941b0 0 185 760 2024-05-01T00:09:18Z Ben 2 Created page with "'''Universal Manipulation Interface''' is a data collection and policy learning framework that allows direct skill transfer from in-the-wild human demonstrations to deployable..." wikitext text/x-wiki '''Universal Manipulation Interface''' is a data collection and policy learning framework that allows direct skill transfer from in-the-wild human demonstrations to deployable robot policies. {{infobox | name = UMI | full_name = Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots | arxiv_link = https://arxiv.org/abs/2402.10329 | project_page = https://umi-gripper.github.io/ | twitter_link = https://twitter.com/chichengcc/status/1758539728444629158 | date = February 2024 | authors = Cheng Chi, Zhenjia xu, Chuer Pan, Eric Cousineau, Benjamin Burchfiel, Siyuan Feng, Russ Tedrake, Shuran Song }} d4365aba0d074505f2eca573af5dbe87b4b4bd9a 761 760 2024-05-01T00:09:25Z Ben 2 wikitext text/x-wiki '''Universal Manipulation Interface''' is a data collection and policy learning framework that allows direct skill transfer from in-the-wild human demonstrations to deployable robot policies. {{infobox paper | name = UMI | full_name = Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots | arxiv_link = https://arxiv.org/abs/2402.10329 | project_page = https://umi-gripper.github.io/ | twitter_link = https://twitter.com/chichengcc/status/1758539728444629158 | date = February 2024 | authors = Cheng Chi, Zhenjia xu, Chuer Pan, Eric Cousineau, Benjamin Burchfiel, Siyuan Feng, Russ Tedrake, Shuran Song }} 511f4975469f1a0914e14a143bff5201b5913bc1 762 761 2024-05-01T00:17:18Z Ben 2 wikitext text/x-wiki '''Universal Manipulation Interface''' is a data collection and policy learning framework that allows direct skill transfer from in-the-wild human demonstrations to deployable robot policies. {{infobox paper | name = UMI | full_name = Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots | arxiv_link = https://arxiv.org/abs/2402.10329 | project_page = https://umi-gripper.github.io/ | twitter_link = https://twitter.com/chichengcc/status/1758539728444629158 | date = February 2024 | authors = Cheng Chi, Zhenjia Xu, Chuer Pan, Eric Cousineau, Benjamin Burchfiel, Siyuan Feng, Russ Tedrake, Shuran Song }} The UMI paper was novel for several reasons: # It completely avoids robot teleoperation and the associated latency. This lets the robot do things like reliably tossing balls. # It provides a low-cost, scalable way to collect lots of data in the wild fbf7abcba7902a7a29661d110d9e323e9465e125 763 762 2024-05-01T00:18:03Z Ben 2 wikitext text/x-wiki '''Universal Manipulation Interface''' is a data collection and policy learning framework that allows direct skill transfer from in-the-wild human demonstrations to deployable robot policies. {{infobox paper | name = UMI | full_name = Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots | arxiv_link = https://arxiv.org/abs/2402.10329 | project_page = https://umi-gripper.github.io/ | twitter_link = https://twitter.com/chichengcc/status/1758539728444629158 | date = February 2024 | authors = Cheng Chi, Zhenjia Xu, Chuer Pan, Eric Cousineau, Benjamin Burchfiel, Siyuan Feng, Russ Tedrake, Shuran Song }} The UMI paper was novel for several reasons: # It completely avoids robot teleoperation and the associated latency. This lets the robot do things like reliably tossing balls. # It provides a low-cost, scalable way to collect lots of data in the wild [[Category: Papers]] 695e305593ccc6074b379f3f3dde624d0c47186e 14 186 764 2024-05-01T00:18:25Z Ben 2 Created page with "This category is used for academic papers related to humanoid robots." wikitext text/x-wiki This category is used for academic papers related to humanoid robots. c97260db0a996c1061150360992d70590f5191b5 0 187 765 2024-05-01T00:55:23Z Ben 2 Created page with "This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN * Power regu..." wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN * Power regulation * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] * Ethernet (to communicate with the head and for debugging) 4d1786526ad8b41f4187c652e6b19b9cb73f6a70 766 765 2024-05-01T01:14:20Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515 ** * Power regulation * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] * Ethernet (to communicate with the head and for debugging) [[Category: Electronics]] 31058132f05946803c66c65995971ef6de835970 767 766 2024-05-01T01:17:30Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranciever * Power regulation * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] * Ethernet (to communicate with the head and for debugging) [[Category: Electronics]] 8b877c02270d388a507df0e516a9bab3fe0bd453 774 767 2024-05-01T01:36:44Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranciever * Power regulation * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] * Ethernet (to communicate with the head and for debugging) [[File:Raspberry Pi CAN hat.jpg|thumb|left]] [[Category: Electronics]] bbf1bcc7fd0caf1bcb19f2da5160367b91533679 775 774 2024-05-01T01:36:57Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranciever * Power regulation * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] * Ethernet (to communicate with the head and for debugging) [[File:Raspberry Pi CAN hat.jpg|thumb|left|Raspberry Pi CAN Hat]] [[Category: Electronics]] ed4b6b6df1a83b2dbace156e1b9f1d404f888399 776 775 2024-05-01T01:37:51Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranciever * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] * Ethernet (to communicate with the head and for debugging) [[File:Raspberry Pi CAN hat.jpg|thumb|left|Raspberry Pi CAN Hat]] [[Category: Electronics]] 231263330d539892c730ec630ff857f81e237854 777 776 2024-05-01T01:41:57Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranciever * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) [[File:Raspberry Pi CAN hat.jpg|thumb|left|Raspberry Pi CAN Hat]] [[Category: Electronics]] 88b47a3f97de190a988bd72bf16ede9ece7bcfd5 778 777 2024-05-01T01:44:38Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranceiver ** MCP2551: CAN Tranceiver * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) [[File:Raspberry Pi CAN hat.jpg|thumb|left|Raspberry Pi CAN Hat]] [[Category: Electronics]] 0bf764d8b7c6cf1d4c844f0b6926a63ce7f4c587 779 778 2024-05-01T01:45:24Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranceiver ** MCP2551: CAN Tranceiver<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) [[File:Raspberry Pi CAN hat.jpg|thumb|left|Raspberry Pi CAN Hat]] === References === <references/> [[Category: Electronics]] bba0a559b20667c1f09e74a351b040040485efa8 780 779 2024-05-01T01:45:33Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy [[File:Raspberry Pi CAN hat.jpg|thumb|left|Raspberry Pi CAN Hat]] == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranceiver ** MCP2551: CAN Tranceiver<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] 85275db397ca95a8bf9ce2ac49b10e3dd920fee9 781 780 2024-05-01T01:46:15Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranceiver ** MCP2551: CAN Tranceiver<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] a70a4558f2f9fcae69a5ed63d6f62b270604c707 6 188 773 2024-05-01T01:36:34Z Ben 2 wikitext text/x-wiki Raspberry Pi CAN hat 363b61b40b643be7cac4b1095a12ae40d85843fa 0 187 782 781 2024-05-01T01:46:23Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranceiver ** MCP2551: CAN Tranceiver<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] === References === <references/> [[Category: Electronics]] 7030c17752a251387c4cc6bef0d98e2ae180e5f5 783 782 2024-05-01T01:46:29Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranceiver ** MCP2551: CAN Tranceiver<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] 3cf932315c7ea228c9660a837cf1999a213fa1b9 784 783 2024-05-01T03:48:20Z Matt 16 Add CAN Transceivers w links wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Tranceiver ** MCP2551: CAN Tranceiver<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] de8661e49e7dd77f8f42f05b888835c0e6f64eeb 785 784 2024-05-01T03:51:45Z Matt 16 Add voltages and links wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * CAN ** MCP2515: CAN Controller, SPI compatible ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] 64c98648b814178a8cac86b711412d3f05535a94 786 785 2024-05-01T03:57:44Z Matt 16 Add MCP2515 link and voltage ranges wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] bbd8058a35f1acb349286fef36763fd8d9378043 787 786 2024-05-01T03:59:06Z Matt 16 Add compute module sections wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Compute Module * Audio * Video * Ethernet (to communicate with the body and for debugging) == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] c49ade2b822aa3cf5a621d3eccd5311dd3d0cf67 788 787 2024-05-01T04:26:03Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Multiplexer * GMSL transmission ** MAX96701 serializer ** MAX96700 deserializer == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] 1129e9859cbfe49b005766ed36910a53ec9d16b4 789 788 2024-05-01T04:28:27Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Multiplexer * GMSL transmission ** MAX96701 serializer ** MAX96700 deserializer ** MAX9295D dual 4-Lane MIPI CSI-2 to GMSL == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] e605c052f57e043f43202c1f478a9e309c408cb5 790 789 2024-05-01T04:28:44Z Ben 2 wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio * Video * Multiplexer * GMSL transmission ** MAX96701 serializer ** MAX96700 deserializer ** MAX9295D dual 4-Lane MIPI CSI-2 to GMSL<ref>https://www.analog.com/en/products/max9295d.html</ref> == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] 5fb4efee4815a2bb5d1fe0b5811fec9039545794 792 790 2024-05-01T05:34:48Z Matt 16 Add Audio Information wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio ** DSP and Audio Codec *** Analog Devices ADAU1452 <ref>https://www.lcsc.com/product-detail/Audio-Interface-ICs_Analog-Devices-ADAU1452KCPZ_C468504.html</ref> ** Amplifier *** Texas Instruments TPA3116D2 <ref>https://www.lcsc.com/product-detail/Audio-Power-OpAmps_Texas-Instruments-TPA3116D2DADR_C50144.html</ref> ** Microphone *** Infineon IM69D130 <ref>https://www.lcsc.com/product-detail/MEMS-Microphones_Infineon-Technologies-IM69D130V01_C536262.html</ref> **Speaker *** FaitalPRO 3FE25 3" Full-Range Speaker Driver<ref>https://faitalpro.com/en/products/LF_Loudspeakers/product_details/index.php?id=401000150</ref> *** * Video * Multiplexer * GMSL transmission ** MAX96701 serializer ** MAX96700 deserializer ** MAX9295D dual 4-Lane MIPI CSI-2 to GMSL<ref>https://www.analog.com/en/products/max9295d.html</ref> == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] a696d94a49e8afbc59d86511152fa6c354445e1d 794 792 2024-05-01T06:30:30Z Matt 16 Add more chips wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio ** DSP and Audio Codec *** Analog Devices ADAU1452 <ref>https://www.lcsc.com/product-detail/Audio-Interface-ICs_Analog-Devices-ADAU1452KCPZ_C468504.html</ref> ** Digital to Analog Converter *** Texas Instruments PCM5102A <ref> https://www.lcsc.com/product-detail/Digital-To-Analog-Converters-DACs_Texas-Instruments-Texas-Instruments-PCM5102APWR_C107671.html </ref> ** Amplifier *** Texas Instruments TPA3116D2 <ref>https://www.lcsc.com/product-detail/Audio-Power-OpAmps_Texas-Instruments-TPA3116D2DADR_C50144.html</ref> ** Microphone *** 2x Infineon IM69D130 <ref>https://www.lcsc.com/product-detail/MEMS-Microphones_Infineon-Technologies-IM69D130V01_C536262.html</ref> **Speaker *** FaitalPRO 3FE25 3" Full-Range Speaker Driver<ref>https://faitalpro.com/en/products/LF_Loudspeakers/product_details/index.php?id=401000150</ref> *** * Video ** 2x MIPI CSI-2 Cameras * Multiplexer ** Video *** MAX9286 *** * GMSL transmission ** MAX96701 serializer ** MAX96700 deserializer ** MAX9295D dual 4-Lane MIPI CSI-2 to GMSL<ref>https://www.analog.com/en/products/max9295d.html</ref> == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] 1e2a27be83b54e7612f51a998a4cfa35163a2703 795 794 2024-05-01T06:46:29Z Matt 16 /* Head */ wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio ** DSP and Audio Codec *** Analog Devices ADAU1452 <ref>https://www.lcsc.com/product-detail/Audio-Interface-ICs_Analog-Devices-ADAU1452KCPZ_C468504.html</ref> ** Digital to Analog Converter *** Texas Instruments PCM5102A <ref> https://www.lcsc.com/product-detail/Digital-To-Analog-Converters-DACs_Texas-Instruments-Texas-Instruments-PCM5102APWR_C107671.html </ref> ** Amplifier *** Texas Instruments TPA3116D2 <ref>https://www.lcsc.com/product-detail/Audio-Power-OpAmps_Texas-Instruments-TPA3116D2DADR_C50144.html</ref> ** Microphone *** 2x Infineon IM69D130 <ref>https://www.lcsc.com/product-detail/MEMS-Microphones_Infineon-Technologies-IM69D130V01_C536262.html</ref> **Speaker *** FaitalPRO 3FE25 3" Full-Range Speaker Driver<ref>https://faitalpro.com/en/products/LF_Loudspeakers/product_details/index.php?id=401000150</ref> *** -- Need audio cable connection -- Need Power connection * Video ** 2x MIPI CSI-2 Cameras * Multiplexer ** Video *** MAX9286? *** * GMSL transmission ** MAX96701 serializer ** MAX96700 deserializer ** MAX9295D dual 4-Lane MIPI CSI-2 to GMSL<ref>https://www.analog.com/en/products/max9295d.html</ref> == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] dd928b3a3f935d2c06f17d2f9ffea93311c4bb9a 796 795 2024-05-01T06:51:28Z Matt 16 add needed parts wikitext text/x-wiki This document describes the PCBs that we use in Stompy == Head == * Audio ** DSP and Audio Codec *** Analog Devices ADAU1452 <ref>https://www.lcsc.com/product-detail/Audio-Interface-ICs_Analog-Devices-ADAU1452KCPZ_C468504.html</ref> ** Digital to Analog Converter *** Texas Instruments PCM5102A <ref> https://www.lcsc.com/product-detail/Digital-To-Analog-Converters-DACs_Texas-Instruments-Texas-Instruments-PCM5102APWR_C107671.html </ref> ** Amplifier *** Texas Instruments TPA3116D2 <ref>https://www.lcsc.com/product-detail/Audio-Power-OpAmps_Texas-Instruments-TPA3116D2DADR_C50144.html</ref> ** Microphone *** 2x Infineon IM69D130 <ref>https://www.lcsc.com/product-detail/MEMS-Microphones_Infineon-Technologies-IM69D130V01_C536262.html</ref> **Speaker *** FaitalPRO 3FE25 3" Full-Range Speaker Driver<ref>https://faitalpro.com/en/products/LF_Loudspeakers/product_details/index.php?id=401000150</ref> *** -- Need audio cable connection -- Need Power connection -- Need SPI Connections * Video ** 2x MIPI CSI-2 Cameras * Multiplexer ** Video *** MAX9286? *** * GMSL transmission ** MAX96701 serializer ** MAX96700 deserializer ** MAX9295D dual 4-Lane MIPI CSI-2 to GMSL<ref>https://www.analog.com/en/products/max9295d.html</ref> == Body == [[File:Raspberry Pi CAN hat.jpg|thumb|right|Raspberry Pi CAN Hat]] * Compute Module * CAN ** MCP2515: CAN Controller, SPI compatible (2.7V~5.5V) <ref>https://www.lcsc.com/product-detail/CAN_MICROCHIP_MCP2515-I-SO_MCP2515-I-SO_C12368.html</ref> ** ATA6561: CAN Transceiver (4.5V~5.5V)<ref> https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-ATA6561-GAQW-N_C616016.html</ref> ** MCP2551: CAN Transceiver (4.5V~5.5V)<ref>https://www.seeedstudio.com/I2C-CAN-Bus-Module-p-5054.html</ref><ref>https://www.lcsc.com/product-detail/CAN-ICs_Microchip-Tech-MCP2551-I-SN_C7376.html</ref> ** TLE6250G: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN_Infineon_TLE6250G_TLE6250G_C111030.html </ref> ** TJA1051: CAN Transceiver (4.5V~5.5V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_NXP-Semicon-TJA1051TK-3-118_C124020.html </ref> ** SN65HVD230: CAN transceiver (3V~3.6V)<ref>https://www.lcsc.com/product-detail/CAN-ICs_Texas-Instruments-SN65HVD230QDR_C16468.html</ref> * Power regulation ** AMS 1117 * IMU ** [https://ozzmaker.com/product/berrygps-imu/ BerryGPS-IMU V4] *** Gyroscope and Accelerometer: LSM6DSL, compatible with SPI and I2C *** Magnetometer: LIS3MDL, compatible with SPI and I2C * Ethernet (to communicate with the head and for debugging) === References === <references/> [[Category: Electronics]] 0b6fe4795c8f44a694a779580b0b7c36e8ca085a 0 189 791 2024-05-01T05:27:56Z Ben 2 Created page with "This page provides some notes on how to apply for SBIR grants. [[Category: Stompy, Expand!]]" wikitext text/x-wiki This page provides some notes on how to apply for SBIR grants. [[Category: Stompy, Expand!]] 1ab18ff9c137e0ca0bbbcdb7040c7e574eb65f75 793 791 2024-05-01T06:04:11Z Stompy 14 Bot expanded article wikitext text/x-wiki == Applying for SBIR Grants == The Small Business Innovation Research (SBIR) program is a United States government program, coordinated by the Small Business Administration (SBA), intended to assist small businesses to engage in Federal Research/Research and Development (R/R&D) activities. The program aims to enhance both the participation of small businesses in Federal R/R&D and the commercialization potentials of their R&D efforts.<ref name="SBA">[https://www.sba.gov/funding-programs/research-grants/small-business-innovation-research-program Small Business Administration: Small Business Innovation Research Program]</ref>. === Eligibility Criteria === To qualify for SBIR grants, applicants should meet several criteria: * Be a for-profit organization. * Have fewer than 500 employees, including affiliates. * Be independently owned and operated, not dominated in its field nationwide. * The principal researcher must be employed by the small business. * The company must be located in the United States. * The small business must be majority U.S. owned by individuals and independently operated.<ref name="SBIR_Overview">[https://www.sbir.gov/overview Small Business Innovation Research (SBIR) Overview]</ref>. === Grant Application Process === 1. Identification of an SBIR funding opportunity at one of the federal agencies that participate in the SBIR program. 2. Registration with the System for Award Management (SAM), SBA Company registry, and the SBIR-specific registration site (e.g., eRA Commons for NIH). 3. Submission of a detailed proposal in response to a specific agency's SBIR opportunity, following the guidelines provided by the respective agency. 4. Peer Review of the submitted proposal. 5. Awarding of the SBIR grant and execution of the proposed R/R&D plan.<ref name="SBAApplication">[https://www.sba.gov/federal-contracting/contracting-guide/small-business-innovation-research-and-small-business-technology-transfer Small Business Administration: Small Business Innovation Research and Small Business Technology Transfer]</ref>. {{infobox grant | name = Small Business Innovation Research (SBIR) Grants | administered_by = Small Business Administration (SBA) | eligibility_link = https://www.sbir.gov/eligibility | application_process_link = https://www.sba.gov/federal-contracting/contracting-guide/small-business-innovation-research-and-small-business-technology-transfer | grant_amount_info = Varies by grant solicitation and agency | catchment_area = United States }} [[Category: Business Funding]] [[Category: Small and Medium Businesses]] [[Category: Innovation]] == References == <references /> f50857be2879e10c73621c799e29af391c2e7470 797 793 2024-05-02T00:24:01Z Ben 2 wikitext text/x-wiki == Applying for SBIR Grants == The Small Business Innovation Research (SBIR) program is a United States government program, coordinated by the Small Business Administration (SBA), intended to assist small businesses to engage in Federal Research/Research and Development (R/R&D) activities. The program aims to enhance both the participation of small businesses in Federal R/R&D and the commercialization potentials of their R&D efforts.<ref name="SBA">[https://www.sba.gov/funding-programs/research-grants/small-business-innovation-research-program Small Business Administration: Small Business Innovation Research Program]</ref>. === Eligibility Criteria === To qualify for SBIR grants, applicants should meet several criteria: * Be a for-profit organization. * Have fewer than 500 employees, including affiliates. * Be independently owned and operated, not dominated in its field nationwide. * The principal researcher must be employed by the small business. * The company must be located in the United States. * The small business must be majority U.S. owned by individuals and independently operated.<ref name="SBIR_Overview">[https://www.sbir.gov/overview Small Business Innovation Research (SBIR) Overview]</ref>. === Grant Application Process === 1. Identification of an SBIR funding opportunity at one of the federal agencies that participate in the SBIR program. 2. Registration with the System for Award Management (SAM), SBA Company registry, and the SBIR-specific registration site (e.g., eRA Commons for NIH). 3. Submission of a detailed proposal in response to a specific agency's SBIR opportunity, following the guidelines provided by the respective agency. 4. Peer Review of the submitted proposal. 5. Awarding of the SBIR grant and execution of the proposed R/R&D plan.<ref name="SBAApplication">[https://www.sba.gov/federal-contracting/contracting-guide/small-business-innovation-research-and-small-business-technology-transfer Small Business Administration: Small Business Innovation Research and Small Business Technology Transfer]</ref>. == References == <references /> 01d800cab727d86ca1899222b4591c52ae2f9971 799 797 2024-05-02T04:35:59Z Ben 2 wikitext text/x-wiki == Applying for SBIR Grants == The Small Business Innovation Research (SBIR) program is a United States government program, coordinated by the Small Business Administration (SBA), intended to assist small businesses to engage in Federal Research/Research and Development (R/R&D) activities. The program aims to enhance both the participation of small businesses in Federal R/R&D and the commercialization potentials of their R&D efforts.<ref name="SBA">[https://www.sba.gov/funding-programs/research-grants/small-business-innovation-research-program Small Business Administration: Small Business Innovation Research Program]</ref>. === Eligibility Criteria === To qualify for SBIR grants, applicants should meet several criteria: * Be a for-profit organization. * Have fewer than 500 employees, including affiliates. * Be independently owned and operated, not dominated in its field nationwide. * The principal researcher must be employed by the small business. * The company must be located in the United States. * The small business must be majority U.S. owned by individuals and independently operated.<ref name="SBIR_Overview">[https://www.sbir.gov/overview Small Business Innovation Research (SBIR) Overview]</ref>. === Grant Application Process === # Identification of an SBIR funding opportunity at one of the federal agencies that participate in the SBIR program. # Registration with the System for Award Management (SAM), SBA Company registry, and the SBIR-specific registration site (e.g., eRA Commons for NIH). # Submission of a detailed proposal in response to a specific agency's SBIR opportunity, following the guidelines provided by the respective agency. # Peer Review of the submitted proposal. # Awarding of the SBIR grant and execution of the proposed R/R&D plan.<ref name="SBAApplication">[https://www.sba.gov/federal-contracting/contracting-guide/small-business-innovation-research-and-small-business-technology-transfer Small Business Administration: Small Business Innovation Research and Small Business Technology Transfer]</ref>. == References == <references /> ce4eedfca58f25ad287850ebffbc0ee0665f54ea 0 61 798 394 2024-05-02T00:44:49Z 108.211.178.220 0 wikitext text/x-wiki This page contains information about how to build a good open-source servo. === Open Source Servos === * [https://github.com/unhuman-io/obot OBot] * [https://github.com/atopile/spin-servo-drive SPIN Servo Drive] * [https://github.com/jcchurch13/Mechaduino-Firmware Mechaduino-Firmware] === Commercial Servos === * [[MyActuator X-Series]] === Controllers === * [https://github.com/napowderly/mcp2515 MCP2515 tranceiver circuit] * [https://github.com/adamb314/ServoProject ServoProject: making a standard servo 36x more accurate] [[Category: Hardware]] [[Category: Electronics]] [[Category: Actuators]] 5428327e83b3f071efee3098db53dfade061665a 0 190 800 2024-05-02T09:35:16Z 14.142.236.26 0 Created page with "[[File:Https://assets-global.website-files.com/60a8e50573e2d780c83ef782/62bf03c35a8471fbacf98b64 Picture1-p-500.png|thumb|Hyperspawn Logo]] Hyperspawn Robotics builds intellig..." wikitext text/x-wiki [[File:Https://assets-global.website-files.com/60a8e50573e2d780c83ef782/62bf03c35a8471fbacf98b64 Picture1-p-500.png|thumb|Hyperspawn Logo]] Hyperspawn Robotics builds intelligent humanoid robots that can operate in two modes: autonomous and teleoperated; allowing the humanoids to function independently or be controlled remotely by a user immersively. The robots use advanced AI and vision LLMs to navigate, make decisions, and execute tasks independently. When in teleoperated mode, users control the robot via VR gear and a motion-tracking suit, providing real-time, intuitive control and enabling a physical presence in inaccessible or hazardous locations. {{infobox company | name = Hyperspawn Robotics | website_link = [https://hyperspawn.co hyperspawn.co] | robots = [[Shadow-1]] }} === Overview === Hyperspawn's humanoid teleoperation system is a reflection of human agility and intelligence, designed to operate in hazardous environments where human intervention is risky. This humanoid robot, controlled via a motion-tracking suit & VR headset, accurately mimics human actions, ensuring precise, and efficient operations in high-risk zones in real time. The design is a result of deep research, collaboration, and a passion for leveraging technology to elevate human lives. === Shadow-1 === [[Shadow-1]], our flagship humanoid robot, is designed to seamlessly integrate into human workspaces. It has been developed with advanced sensor technology and AI-driven functionalities, enabling it to perform tasks ranging from simple material handling to complex problem-solving scenarios that require interaction with human counterparts. Shadow-1 is designed for both autonomous and teleoperated modes. Equipped with advanced sensors and actuators, it can navigate complex environments and perform detailed tasks. Its teleoperation capabilities, allows users to control the robot remotely with high precision and real-time feedback, effectively placing them anywhere Shadow-1 can go. This makes it ideal for applications in hazardous conditions, remote assistance, and scenarios requiring human-like interaction without the risks. === Global Impact === Hyperspawn's intent is to have more humanoids in the world for better real world experience for humans. It will be the first real, interactive and fluid device which has the capability of taking physical actions on your behalf and having a personality and aligning to yours well. Teleoperation feature is the best way to enter peoples life as other people ( physical video calls) and will lead to more and better social connections and even be very useful in B2B applications like Manufacturing Facilities, Defense, etc. === Future Vision === Hyperspawn Robotics envisions a world with infinite mobility for humanity. Where humans can physically interact from anywhere in the world without the constraints of space and time. Where tasks in extreme conditions, be it in defense, healthcare, space exploration, etc., are seamlessly executed by humanoid robots, ensuring not only human safety but also guaranteed efficiency. And, this is just an innovative stepping stone in our journey to develop the most advanced autonomous humanoids, harnessing the latest advances in AI and integrating them seamlessly into robotics. Sounds too futuristic? We’re making it happen! [[Category:Companies]] 4095901ab0c9d4e4cdc25dce0df06c8268caea98 802 800 2024-05-02T09:39:37Z Evolve 20 wikitext text/x-wiki [[File:HyperspawnRoboticsLogo.png|thumb|150px|Hyperspawn Robotics]] Hyperspawn Robotics builds intelligent humanoid robots that can operate in two modes: autonomous and teleoperated; allowing the humanoids to function independently or be controlled remotely by a user immersively. The robots use advanced AI and vision LLMs to navigate, make decisions, and execute tasks independently. When in teleoperated mode, users control the robot via VR gear and a motion-tracking suit, providing real-time, intuitive control and enabling a physical presence in inaccessible or hazardous locations. {{infobox company | name = Hyperspawn Robotics | website_link = [https://hyperspawn.co hyperspawn.co] | robots = [[Shadow-1]] }} === Overview === Hyperspawn's humanoid teleoperation system is a reflection of human agility and intelligence, designed to operate in hazardous environments where human intervention is risky. This humanoid robot, controlled via a motion-tracking suit & VR headset, accurately mimics human actions, ensuring precise, and efficient operations in high-risk zones in real time. The design is a result of deep research, collaboration, and a passion for leveraging technology to elevate human lives. === Shadow-1 === [[Shadow-1]], our flagship humanoid robot, is designed to seamlessly integrate into human workspaces. It has been developed with advanced sensor technology and AI-driven functionalities, enabling it to perform tasks ranging from simple material handling to complex problem-solving scenarios that require interaction with human counterparts. Shadow-1 is designed for both autonomous and teleoperated modes. Equipped with advanced sensors and actuators, it can navigate complex environments and perform detailed tasks. Its teleoperation capabilities, allows users to control the robot remotely with high precision and real-time feedback, effectively placing them anywhere Shadow-1 can go. This makes it ideal for applications in hazardous conditions, remote assistance, and scenarios requiring human-like interaction without the risks. === Global Impact === Hyperspawn's intent is to have more humanoids in the world for better real world experience for humans. It will be the first real, interactive and fluid device which has the capability of taking physical actions on your behalf and having a personality and aligning to yours well. Teleoperation feature is the best way to enter peoples life as other people ( physical video calls) and will lead to more and better social connections and even be very useful in B2B applications like Manufacturing Facilities, Defense, etc. === Future Vision === Hyperspawn Robotics envisions a world with infinite mobility for humanity. Where humans can physically interact from anywhere in the world without the constraints of space and time. Where tasks in extreme conditions, be it in defense, healthcare, space exploration, etc., are seamlessly executed by humanoid robots, ensuring not only human safety but also guaranteed efficiency. And, this is just an innovative stepping stone in our journey to develop the most advanced autonomous humanoids, harnessing the latest advances in AI and integrating them seamlessly into robotics. Sounds too futuristic? We’re making it happen! [[Category:Companies]] 5d035545551679292d2c657839a530c6d42c4c7a 805 802 2024-05-02T21:54:15Z Ben 2 wikitext text/x-wiki [[File:HyperspawnRoboticsLogo.png|thumb|150px|Hyperspawn Robotics]] Hyperspawn Robotics builds intelligent humanoid robots that can operate in two modes: autonomous and teleoperated; allowing the humanoids to function independently or be controlled remotely by a user immersively. The robots use advanced AI and vision LLMs to navigate, make decisions, and execute tasks independently. When in teleoperated mode, users control the robot via VR gear and a motion-tracking suit, providing real-time, intuitive control and enabling a physical presence in inaccessible or hazardous locations. {{infobox company | name = Hyperspawn Robotics | website_link = https://hyperspawn.co | robots = [[Shadow-1]] }} === Overview === Hyperspawn's humanoid teleoperation system is a reflection of human agility and intelligence, designed to operate in hazardous environments where human intervention is risky. This humanoid robot, controlled via a motion-tracking suit & VR headset, accurately mimics human actions, ensuring precise, and efficient operations in high-risk zones in real time. The design is a result of deep research, collaboration, and a passion for leveraging technology to elevate human lives. === Shadow-1 === [[Shadow-1]], our flagship humanoid robot, is designed to seamlessly integrate into human workspaces. It has been developed with advanced sensor technology and AI-driven functionalities, enabling it to perform tasks ranging from simple material handling to complex problem-solving scenarios that require interaction with human counterparts. Shadow-1 is designed for both autonomous and teleoperated modes. Equipped with advanced sensors and actuators, it can navigate complex environments and perform detailed tasks. Its teleoperation capabilities, allows users to control the robot remotely with high precision and real-time feedback, effectively placing them anywhere Shadow-1 can go. This makes it ideal for applications in hazardous conditions, remote assistance, and scenarios requiring human-like interaction without the risks. === Global Impact === Hyperspawn's intent is to have more humanoids in the world for better real world experience for humans. It will be the first real, interactive and fluid device which has the capability of taking physical actions on your behalf and having a personality and aligning to yours well. Teleoperation feature is the best way to enter peoples life as other people ( physical video calls) and will lead to more and better social connections and even be very useful in B2B applications like Manufacturing Facilities, Defense, etc. === Future Vision === Hyperspawn Robotics envisions a world with infinite mobility for humanity. Where humans can physically interact from anywhere in the world without the constraints of space and time. Where tasks in extreme conditions, be it in defense, healthcare, space exploration, etc., are seamlessly executed by humanoid robots, ensuring not only human safety but also guaranteed efficiency. And, this is just an innovative stepping stone in our journey to develop the most advanced autonomous humanoids, harnessing the latest advances in AI and integrating them seamlessly into robotics. Sounds too futuristic? We’re making it happen! [[Category:Companies]] 4d95b42bf03a3d121274ad1e5ccb73233847d32c 6 191 801 2024-05-02T09:37:34Z Evolve 20 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 1 803 700 2024-05-02T18:46:13Z Jos 17 /* Getting Started */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === Here are some resources to get started learning about humanoid robots. {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 77779a547fddb8bad0a1ff053ccfd2185cb516a3 806 803 2024-05-03T15:50:11Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 2fcc461504a84faeb5c1207692973d5f843026ed 810 806 2024-05-03T17:10:54Z 129.97.185.73 0 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 01a22aa90f0cf5977d1dd7b5433de6ae22965bb4 815 810 2024-05-03T17:26:48Z 129.97.185.73 0 /* List of Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | An actuator built for the DEEP Robotics quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 5db9be535742a2cb413bfe24c1c2102afcdafeac 817 815 2024-05-03T17:41:13Z 129.97.185.73 0 /* List of Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the DEEP Robotics quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} aad57f7db9598578f42bc74471a72508c541e993 819 817 2024-05-03T17:45:52Z 129.97.185.73 0 /* List of Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |} 9930eefa069f44676308fdeba0b359488e179218 14 192 804 2024-05-02T18:47:25Z Jos 17 Created page with "== Open Source == * https://open-dynamic-robot-initiative.github.io/" wikitext text/x-wiki == Open Source == * https://open-dynamic-robot-initiative.github.io/ 4ed9c587ac0752ad133771aa0c474d12420a5a08 0 193 807 2024-05-03T15:50:51Z Ben 2 Created page with "This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it!" wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! ff4b4af494c5d0c817cc65c2fed3dcbc2a613cd2 820 807 2024-05-03T20:13:40Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' '''Getting Started with Building and Experimenting with Your Own Humanoid Robot ''' '''Introduction''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. Key Components and Tools Fundamentals to Get You Started: Humanoid Robot Anatomy: Understand the basics of sensors, actuators, and controllers. Consider what makes a humanoid robot function—from the planetary gear configurations in the joints to the complex sensor arrays for environmental interaction. Simulation Tools: ISAAC Sim by NVIDIA: Before assembling your robot, simulate your designs in ISAAC Sim. This tool is perfect for experimenting with different configurations in a controlled virtual environment, minimizing costs and potential damage to physical components. Building Your Humanoid Robot Selecting Components: Actuators and Gearboxes: Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. Assembly Tips: Community Forums: Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. Programming and Control ROS (Robot Operating System): Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. Custom Software Solutions: Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. Experimenting with Your Humanoid Robot Testing and Iteration: Virtual before Physical: Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. Real-World Testing: Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. Data Collection and Analysis: Camera Systems: Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. Advanced Customization and Community Engagement Open Source Projects: Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. Modular Design: Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. Safety, Ethics, and Continuous Learning Safety Protocols: Always implement robust safety measures when testing and demonstrating your robot. f698531ee5696770ea613c467c9ed371902cf4d1 821 820 2024-05-03T20:19:28Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' '''Getting Started with Building and Experimenting with Your Own Humanoid Robot This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == '''Key Components and Tools''' == ===== Fundamentals to Get You Started ===== Humanoid Robot Anatomy: Understand the basics of sensors, actuators, and controllers. Consider what makes a humanoid robot function—from the planetary gear configurations in the joints to the complex sensor arrays for environmental interaction. ===== Simulation Tools: ===== ISAAC Sim by NVIDIA: Before assembling your robot, simulate your designs in ISAAC Sim. This tool is perfect for experimenting with different configurations in a controlled virtual environment, minimizing costs and potential damage to physical components. == Building Your Humanoid Robot == Selecting Components: Actuators and Gearboxes: Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. Assembly Tips: Community Forums: Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. Programming and Control ROS (Robot Operating System): Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. Custom Software Solutions: Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. Experimenting with Your Humanoid Robot Testing and Iteration: Virtual before Physical: Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. Real-World Testing: Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. Data Collection and Analysis: Camera Systems: Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. Advanced Customization and Community Engagement Open Source Projects: Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. Modular Design: Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. Safety, Ethics, and Continuous Learning Safety Protocols: Always implement robust safety measures when testing and demonstrating your robot. e9964219991d1e74022ffef7ba64e97982d5e7e1 822 821 2024-05-03T20:21:58Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' '''Getting Started with Building and Experimenting with Your Own Humanoid Robot This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Key Components and Tools == ===== Fundamentals to Get You Started ===== Humanoid Robot Anatomy: Understand the basics of sensors, actuators, and controllers. Consider what makes a humanoid robot function—from the planetary gear configurations in the joints to the complex sensor arrays for environmental interaction. ===== Simulation Tools: ===== ISAAC Sim by NVIDIA: Before assembling your robot, simulate your designs in ISAAC Sim. This tool is perfect for experimenting with different configurations in a controlled virtual environment, minimizing costs and potential damage to physical components. == Building Your Humanoid Robot == Selecting Components: Actuators and Gearboxes: Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. Assembly Tips: Community Forums: Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. Programming and Control ROS (Robot Operating System): Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. Custom Software Solutions: Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. Experimenting with Your Humanoid Robot Testing and Iteration: Virtual before Physical: Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. Real-World Testing: Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. Data Collection and Analysis: Camera Systems: Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. Advanced Customization and Community Engagement Open Source Projects: Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. Modular Design: Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. Safety, Ethics, and Continuous Learning Safety Protocols: Always implement robust safety measures when testing and demonstrating your robot. ad2ddd376428b10b34acfdd736e699caa2857810 823 822 2024-05-03T20:30:10Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Key Components and Tools == ===== Fundamentals to Get You Started ===== Humanoid Robot Anatomy: Understand the basics of sensors, actuators, and controllers. Consider what makes a humanoid robot function—from the planetary gear configurations in the joints to the complex sensor arrays for environmental interaction. ===== Simulation Tools: ===== ISAAC Sim by NVIDIA: Before assembling your robot, simulate your designs in ISAAC Sim. This tool is perfect for experimenting with different configurations in a controlled virtual environment, minimizing costs and potential damage to physical components. == Building Your Humanoid Robot == ===== '''Selecting Components'' ===== ' Actuators and Gearboxes: Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. Assembly Tips: Community Forums: Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. Programming and Control ROS (Robot Operating System): Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. Custom Software Solutions: Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. Experimenting with Your Humanoid Robot Testing and Iteration: Virtual before Physical: Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. Real-World Testing: Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. Data Collection and Analysis: Camera Systems: Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. Advanced Customization and Community Engagement Open Source Projects: Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. Modular Design: Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. Safety, Ethics, and Continuous Learning Safety Protocols: Always implement robust safety measures when testing and demonstrating your robot. a94a0d1f46b6df1b3580b16de2448a96b2d21f7c 824 823 2024-05-03T20:31:25Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Key Components and Tools == ===== Fundamentals to Get You Started ===== Humanoid Robot Anatomy: Understand the basics of sensors, actuators, and controllers. Consider what makes a humanoid robot function—from the planetary gear configurations in the joints to the complex sensor arrays for environmental interaction. ===== Simulation Tools: ===== ISAAC Sim by NVIDIA: Before assembling your robot, simulate your designs in ISAAC Sim. This tool is perfect for experimenting with different configurations in a controlled virtual environment, minimizing costs and potential damage to physical components. == Building Your Humanoid Robot == === Selecting Components === ==== Actuators and Gearboxes ==== Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. Assembly Tips: Community Forums: Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. Programming and Control ROS (Robot Operating System): Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. Custom Software Solutions: Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. Experimenting with Your Humanoid Robot Testing and Iteration: Virtual before Physical: Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. Real-World Testing: Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. Data Collection and Analysis: Camera Systems: Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. Advanced Customization and Community Engagement Open Source Projects: Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. Modular Design: Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. Safety, Ethics, and Continuous Learning Safety Protocols: Always implement robust safety measures when testing and demonstrating your robot. 1e5d8f0e61cc6378a613c68e7cff3facaf2e16a5 825 824 2024-05-03T20:38:17Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == === Selecting Components === ==== Actuators and Gearboxes ==== Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. ==== Assembly Tips ==== ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. ==== Actuators and Gearboxes ==== Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. Assembly Tips: Community Forums: Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. Programming and Control ROS (Robot Operating System): Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. Custom Software Solutions: Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. Experimenting with Your Humanoid Robot Testing and Iteration: Virtual before Physical: Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. Real-World Testing: Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. Data Collection and Analysis: Camera Systems: Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. Advanced Customization and Community Engagement Open Source Projects: Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. Modular Design: Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. Safety, Ethics, and Continuous Learning Safety Protocols: Always implement robust safety measures when testing and demonstrating your robot. 5a4bcff2ef6663f92db2634665e64b5bea7e4d3a 826 825 2024-05-03T20:38:53Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == === Selecting Components === ==== Actuators and Gearboxes ==== Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. ==== Assembly Tips ==== ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others, such as the Kayra project, a 3D printable open-source humanoid robot. === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 9e8827f79564613254cba3c6074a893e7f1156d1 827 826 2024-05-03T20:41:23Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == === Selecting Components === ==== Actuators and Gearboxes ==== Whether you're looking at traditional servos or exploring advanced options like the MyActuator X-Series or cycloidal gears, understanding the torque, speed, and precision each component offers is crucial. Check out community-generated charts and databases for a breakdown of cost vs. performance. ==== Assembly Tips ==== ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 3feff07d7b536676b065eb293b56a5af0ddb627a 828 827 2024-05-03T21:25:58Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == === Selecting Components === ==== Actuators and Gearboxes ==== == Actuators in Humanoid Robotics: Insights and Innovations == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. === Global Collaborations and Innovations === ==== Actuator Design Finalization for Public Release ==== Collaborations, such as those discussed for finalizing actuator designs for public releases, highlight the global effort in refining robotic components. These collaborations often result in minor but crucial adjustments that enhance the overall functionality and integration of actuators into robotic systems. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 9d8875084327c2dbc3b40f69011298347ea93da5 829 828 2024-05-03T21:27:38Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == === Selecting Components === == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. af5cfcf90d94d5f43eeb78be516f9d765b2c3fc5 830 829 2024-05-03T21:30:02Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 7e5f226b5e36d2e9fe249d8aa1074ed1d761c81f 831 830 2024-05-03T21:34:43Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! update:''work in progress - starting with a template, plan to expand on sections :)'' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 2467a3776ddffb1b0c4f1723c451a0eb62d9dc34 0 16 808 718 2024-05-03T16:55:18Z Ben 2 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. === Onboarding === To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Cluster 1 === === Cluster 2 === The cluster has 8 available nodes (each with 8 GPUs): <syntaxhighlight lang="text"> compute-permanent-node-68 compute-permanent-node-285 compute-permanent-node-493 compute-permanent-node-625 compute-permanent-node-626 compute-permanent-node-749 compute-permanent-node-801 compute-permanent-node-580 </syntaxhighlight> When you ssh-in, you log in to the bastion node pure-caribou-bastion from which you can log in to any other node where you can test your code. == Reserving a GPU == Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> [[Category:K-Scale]] 7ce1b1cb7f0cf12db1938ae0f41fdf5f6c32b895 809 808 2024-05-03T16:56:01Z Ben 2 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Cluster 1 === === Cluster 2 === The cluster has 8 available nodes (each with 8 GPUs): <syntaxhighlight lang="text"> compute-permanent-node-68 compute-permanent-node-285 compute-permanent-node-493 compute-permanent-node-625 compute-permanent-node-626 compute-permanent-node-749 compute-permanent-node-801 compute-permanent-node-580 </syntaxhighlight> When you ssh-in, you log in to the bastion node pure-caribou-bastion from which you can log in to any other node where you can test your code. == Reserving a GPU == Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> [[Category:K-Scale]] 17872a9f6b0bb2914c7b58f28fedb491a1f1c5c9 0 194 811 2024-05-03T17:15:10Z 129.97.185.73 0 Created page with "[https://mirsee.com/ Mirsee Robotics] is a robotics company based in Cambridge, ON, Canada. Along with custom actuators, hands, and other solutions, they have developed two hu..." wikitext text/x-wiki [https://mirsee.com/ Mirsee Robotics] is a robotics company based in Cambridge, ON, Canada. Along with custom actuators, hands, and other solutions, they have developed two humanoid robots, [[Beomni]] and [[Mirsee]], both of which have wheeled bases. {{infobox company | name = Mirsee Robotics | country = Canada | website_link = https://mirsee.com/ | robots = [[Beomni]], [[Mirsee]] }} [[Category:Companies]] 310d7ecb7bac430f83342c26498b2696e3d3c03c 0 195 812 2024-05-03T17:20:54Z 129.97.185.73 0 Created page with "Mirsee is a humanoid robot from [[Mirsee Robotics]]. {{infobox robot | name = Mirsee | organization = [[Mirsee Robotics]] | height | weight | single_hand_payload | two_hand_p..." wikitext text/x-wiki Mirsee is a humanoid robot from [[Mirsee Robotics]]. {{infobox robot | name = Mirsee | organization = [[Mirsee Robotics]] | height | weight | single_hand_payload | two_hand_payload = 60 lbs | video_link = https://www.youtube.com/watch?v=mbHCuPBsIgg | cost }} [[Category:Robots]] f444cf6316711d299d16989f7c5145080092a322 814 812 2024-05-03T17:23:36Z 129.97.185.73 0 wikitext text/x-wiki Beomni is a humanoid robot developed by [[Mirsee Robotics]] for the Beyond Imagination AI company: https://www.beomni.ai/. {{infobox robot | name = Beomni | organization = [[Mirsee Robotics]] | height | weight | single_hand_payload | two_hand_payload | video_link = https://www.youtube.com/watch?v=chIFBTHyEaE | cost }} [[Category:Robots]] eeafd4009e739d39500ceabcdb10b715d40b9b80 0 196 813 2024-05-03T17:21:03Z 129.97.185.73 0 Created page with "Mirsee is a humanoid robot from [[Mirsee Robotics]]. {{infobox robot | name = Mirsee | organization = [[Mirsee Robotics]] | height | weight | single_hand_payload | two_hand_p..." wikitext text/x-wiki Mirsee is a humanoid robot from [[Mirsee Robotics]]. {{infobox robot | name = Mirsee | organization = [[Mirsee Robotics]] | height | weight | single_hand_payload | two_hand_payload = 60 lbs | video_link = https://www.youtube.com/watch?v=mbHCuPBsIgg | cost }} [[Category:Robots]] f444cf6316711d299d16989f7c5145080092a322 0 197 816 2024-05-03T17:40:05Z 129.97.185.73 0 Created page with "The [https://www.deeprobotics.cn/en/index/j60.html J60] by [[DEEP Robotics]] is a family of high-performance actuators designed for use in quadruped and humanoid robots. == O..." wikitext text/x-wiki The [https://www.deeprobotics.cn/en/index/j60.html J60] by [[DEEP Robotics]] is a family of high-performance actuators designed for use in quadruped and humanoid robots. == Overview == The J60 actuators are high-performance actuators have been specifically designed for use in humanoid robots and in [[DEEP Robotics]] quadrupeds. These actuators are reliable and durable, featuring a high torque-to-weight ratio. The joints include a gear reduction, a frameless torque motor, a servo driver, and an absolute value encoder into one compact unit. === Actuators === {{infobox actuator | name = J60-6 | manufacturer = DEEP Robotics | link = https://www.deeprobotics.cn/en/index/j60.html | peak_torque = 19.94 Nm | peak_speed = 24.18 rad/s | dimensions = 76.5mm diameter, 63mm length | weight = 480g | absolute_encoder_resolution = 14 bit | operating_voltage_range = 12-36V | standard_operating_voltage = 24V | interface = CAN bus / RS485 | control_frequency = 1kHz }} {{infobox actuator | name = J60-10 | manufacturer = DEEP Robotics | link = https://www.deeprobotics.cn/en/index/j60.html | peak_torque = 30.50 Nm | peak_speed = 15.49 rad/s | dimensions = 76.5mm diameter, 72.5mm length | weight = 540g | absolute_encoder_resolution = 14 bit | operating_voltage_range = 12-36V | standard_operating_voltage = 24V | interface = CAN bus / RS485 | control_frequency = 1kHz }} [[Category: Actuators]] b5cdee73845d295957f9422b4ae2f30130be176c 0 198 818 2024-05-03T17:45:18Z 129.97.185.73 0 Created page with "[https://www.deeprobotics.cn/en/ DEEP Robotics] is a robotics company based in Hangzhou, Zhejiang Province, China. They make research-grade and industrial-level quadrupedal ro..." wikitext text/x-wiki [https://www.deeprobotics.cn/en/ DEEP Robotics] is a robotics company based in Hangzhou, Zhejiang Province, China. They make research-grade and industrial-level quadrupedal robots, such as the [https://www.deeprobotics.cn/en/index/product1.html Lite 3] and the [https://www.deeprobotics.cn/en/index/product3.html X30], as well as high-performance actuators for humanoid and quadruped robots (the [[J60]] series). {{infobox company | name = DEEP Robotics | country = China | website_link = https://www.deeprobotics.cn/en/ }} [[Category:Companies]] 28ba58276da9ea11e9bd04fc9f44af5e22160df4 0 193 832 831 2024-05-03T21:35:06Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual before Physical ==== Use ISAAC Sim to test your designs under various simulated conditions to refine your robot's mechanics and electronics. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 353d7921ff47c5ccd6125524a4bfa2ae44d78ef0 843 832 2024-05-04T20:00:14Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. - MuJoCo (MJX): Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. - VSim: Claims to be 10x faster than other simulators. - ManiSkill/Sapien: Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. == Best Practices for Virtual Testing == - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: - Use Azure's A100 GPUs for Isaac training. - Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. c9cda6bafed17dbebdb2d645d276f30f8cb0067b 844 843 2024-05-04T20:09:14Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. == Best Practices for Virtual Testing == - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. ==== Real-World Testing ==== Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 4c1720c62fabb5ec7632e59e6f72da89ab422efa 845 844 2024-05-04T20:12:48Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 16cacb5853202dbd117a26819f958e48729f7a5b 846 845 2024-05-04T20:13:18Z Vrtnis 21 /* VSim */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 0b889dab046d6fc518b669a5b02ff4dace2f045a 847 846 2024-05-04T20:14:53Z Vrtnis 21 /* Best Practices for Virtual Testing */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [SPIN project](https://github.com/atopile/spin-servo-drive) by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 2622dcd3902c600f037cd27d18576559f7e9e046 848 847 2024-05-04T20:32:15Z Vrtnis 21 /* SPIN: A Revolutionary Servo Project */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 2b1fa6ab0fdc5962a535197892ca7e3459ba2b91 849 848 2024-05-04T21:35:46Z Vrtnis 21 /* Planetary and Cycloidal Gear Actuators */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 83b4f61e0fd355379385981fb1873046311345a2 850 849 2024-05-04T21:36:17Z Vrtnis 21 /* Planetary and Cycloidal Gear Actuators */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. M yActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. dbba943ce680c422b74ab57a8900f30bc33ebc57 851 850 2024-05-04T21:36:41Z Vrtnis 21 /* Planetary and Cycloidal Gear Actuators */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 2a121f652bc54716a1fe52ea1ac81aa90460f6ea 852 851 2024-05-04T21:49:25Z Vrtnis 21 /* Actuator Types and Design Inspirations */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 788c00bf689b67edc845a0e1a0ca56f7b5a65423 854 852 2024-05-05T06:40:29Z Vrtnis 21 /* ROS (Robot Operating System) */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. ce31111d3fee041997fb0cd72a675bb976e27efd 855 854 2024-05-05T18:43:18Z Vrtnis 21 /* MIT Cheetah Actuator */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. In short, if you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. c2a41f09e5d28f4ff1c775e2152e4de0b900b86c 856 855 2024-05-05T18:43:43Z Vrtnis 21 /* MIT Cheetah Actuator */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. c68c04e769f5357c86f55b1b1798141fb5594940 857 856 2024-05-05T18:45:01Z Vrtnis 21 /* MIT Cheetah Actuator */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. bae0cba8c9b7950a9221c08098444ad4995f30a8 858 857 2024-05-05T18:45:20Z Vrtnis 21 /* MIT Cheetah Actuator */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 743407e54cf5ff832358cf51ce76da1a666a1e35 859 858 2024-05-05T19:03:43Z Vrtnis 21 /* Comprehensive Actuator Comparisons */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== There are discussions about custom actuator developments tailored for specific applications. For example, discussions from [Iris Dynamics on electric linear actuators](https://irisdynamics.com/products/orca-series) suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 9019a2ef9524c334459c6cbc44cfccefbc1e44a1 860 859 2024-05-05T19:04:23Z Vrtnis 21 /* Custom Actuator Developments */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics on electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 5b6e66ccd4c3e3c5d6149cceba71eb1ee9bf4460 861 860 2024-05-05T19:04:42Z Vrtnis 21 /* Custom Actuator Developments */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. fb24b5b43ce639d32511f0282c5cdd102cb68740 862 861 2024-05-05T19:11:08Z Vrtnis 21 /* MuJoCo (MJX) */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. An initial conversion was done for Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. e9fbe284dd5453c5748313d951368ab351fb5e2e 863 862 2024-05-05T19:11:52Z Vrtnis 21 /* MuJoCo (MJX) */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete; you can help by expanding it! '''update:''work in progress - starting with a template, plan to expand on sections :)''''' This guide is crafted for enthusiasts who are not just looking to study humanoid robotics but to actually build and experiment with their own robots. == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 7e7338c7af7aaa0435222cc2bd2012fc43ed525f 864 863 2024-05-05T19:27:47Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 9f8acc36732dec1fea191c0bb0be97e41a224cbd 876 864 2024-05-06T06:18:19Z Vrtnis 21 wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 60e633e04ad9c6dc71f614b955a04484ee964df7 877 876 2024-05-06T06:25:01Z Vrtnis 21 /* Series Elastic and Quasi-Direct Drive Actuators */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some points to consider: * Pick Your Springs Wisely: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between precise torque control and getting sluggish responses. * * Calibrate, Calibrate, Calibrate: Calibration is your best buddy. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep your movements smooth and predictable. * * Control Loops FTW: You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. Go for PID controllers perhaps. * * Friction: Friction can impact your torque control big time, especially in gearboxes and linkages. Low-friction components will help keep everything moving smoothly. * * Load Path Matters: Make sure your spring is right in the direct load path between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing. * * Watch Out for Fatigue: Springs aren’t indestructible. With a lot of high-impact activities, they can wear out. Keep an eye on them to avoid breakdowns at the worst possible moment. * * Get Your Software Right: SEAs thrive on real-time feedback. Ensure your software can handle data fast enough, maybe even use a real-time operating system or optimized signal processing. * ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. b34c16bcd9ca60f69f38061a50e745d2451d6951 878 877 2024-05-06T06:26:37Z Vrtnis 21 /* Series Elastic and Quasi-Direct Drive Actuators */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Finding that sweet spot is key. Calibration is your best buddy. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost. Springs aren’t indestructible. If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. ff070237df72a4ef35fe2813437568acecd49cb5 879 878 2024-05-06T06:26:57Z Vrtnis 21 /* Series Elastic and Quasi-Direct Drive Actuators */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Calibration is your best buddy. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost. Springs aren’t indestructible. If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. b169e5ad51bd9d40c6a8fb37d35ccb54cb397c87 880 879 2024-05-06T06:35:34Z Vrtnis 21 /* Series Elastic and Quasi-Direct Drive Actuators */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost.If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to or start your own open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 5618194ecfcd1e82f4ee4e8138a05febba093427 881 880 2024-05-06T17:43:44Z Vrtnis 21 /* Open Source Projects */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the MIT Cheetah actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost.If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to an open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs. Check out [https://humanoids.wiki/w/Stompy Stompy!] === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 6737df27ab99200f97333623c586542b4ccc883c 882 881 2024-05-06T17:47:18Z Vrtnis 21 /* Building Your Humanoid Robot */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the [https://humanoids.wiki/w/MIT_Cheetah MIT Cheetah] actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost.If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to an open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs. Check out [https://humanoids.wiki/w/Stompy Stompy!] === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 01ba2abedbfdd94f1174f9a5ddd33fc34fefc07b 0 1 833 819 2024-05-03T23:41:43Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |} fc3f5167ec8e3dfd7915f0824bcf6e4207c067e5 871 833 2024-05-06T03:51:49Z Dymaxion 22 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[NASA Johnson Space Center]] | [[Robonaut2]] |} 90d35249297815078c5ad24b12effa6f0079a7e3 873 871 2024-05-06T04:10:32Z Dymaxion 22 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[NASA]] | [[Robonaut2]] |} 8fb84ca6e653bf36575a46741ef9a945f4acda50 875 873 2024-05-06T05:34:57Z Ben 2 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |} b08def286007fc5165e8a0e5c8ee9c4155b5a743 0 199 834 2024-05-03T23:44:29Z Modeless 7 Created page with "SUPCON is a Chinese company that established a lab called the "Zhejiang Humanoid Robot Innovation Center". They have a humanoid called [[Navigator α]].<ref>https://www.therob..." wikitext text/x-wiki SUPCON is a Chinese company that established a lab called the "Zhejiang Humanoid Robot Innovation Center". They have a humanoid called [[Navigator α]].<ref>https://www.therobotreport.com/supcon-opens-innovation-center-launches-navigator-%CE%B1-humanoid-robot/</ref><ref>https://global.supcon.com/news/info/817.html</ref> a8af0823909865bac31302c82052128b5e5c9946 0 5 835 720 2024-05-04T19:00:04Z Ben 2 wikitext text/x-wiki [[File:Logo.png|right|200px|thumb|The K-Scale Labs logo.]] [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} [[Category:Companies]] [[Category:K-Scale]] 6667079512b772b6d5a7953bcc80de0e96066370 840 835 2024-05-04T19:13:53Z Ben 2 wikitext text/x-wiki [[File:Logo.png|right|200px|thumb|The K-Scale Labs logo.]] [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} == Logos == Here are some other versions of the K-Scale Labs logo. <gallery> Logo.png K-Scale Labs Raw Logo.png K-Scale Padded Logo.png K-Scale Raw Padded Logo.png KScale Raw Transparent Padded Logo.png </gallery> === Adding Padding === Here's a helpful command to add padding to an image using ImageMagick: <syntaxhighlight lang="bash"> convert \ -channel RGB \ -negate \ -background black \ -alpha remove \ -alpha off \ -gravity center \ -scale 384x384 \ -extent 512x512 \ logo.png \ logo_padded.png </syntaxhighlight> [[Category:Companies]] [[Category:K-Scale]] 413302d02e1cabb55d1d4df8fd062bfde6bbfd4c 842 840 2024-05-04T19:24:02Z Ben 2 wikitext text/x-wiki [[File:Logo.png|right|200px|thumb|The K-Scale Labs logo.]] [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} == Logos == Here are some other versions of the K-Scale Labs logo. <gallery> Logo.png K-Scale Labs Raw Logo.png K-Scale Padded Logo.png K-Scale Raw Padded Logo.png KScale Raw Transparent Padded Logo.png K-Scale Raw Padded White Logo.png </gallery> === Adding Padding === Here's a helpful command to add padding to an image using ImageMagick: <syntaxhighlight lang="bash"> convert \ -channel RGB \ -negate \ -background black \ -alpha remove \ -alpha off \ -gravity center \ -scale 384x384 \ -extent 512x512 \ logo.png \ logo_padded.png </syntaxhighlight> [[Category:Companies]] [[Category:K-Scale]] 419fc39dd3c3592d3c48e7db6ce0d6e43e5c6c9b 6 200 836 2024-05-04T19:02:11Z Ben 2 wikitext text/x-wiki K-Scale Labs Logo without the K 3e249413ed3ef595115c3b5e88dc00c74032bd96 6 201 837 2024-05-04T19:06:48Z Ben 2 wikitext text/x-wiki K-Scale Padded Logo 0435f8c4095d5512d672fa5c8d9194830f933ead 6 202 838 2024-05-04T19:07:17Z Ben 2 wikitext text/x-wiki K-Scale Raw Padded Logo cef7879158402f833b87763f610a160eca284466 6 203 839 2024-05-04T19:13:32Z Ben 2 wikitext text/x-wiki KScale Raw Transparent Padded Logo d8e27e2c1ef230cec39664f1b563e2948d801aa6 6 204 841 2024-05-04T19:23:52Z Ben 2 wikitext text/x-wiki K-Scale Raw Padded White Logo f56377bd5347514d9f9d6f502da1b23b38efffd6 0 16 853 809 2024-05-05T00:48:20Z Jos 17 /* Reserving a GPU */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Cluster 1 === === Cluster 2 === The cluster has 8 available nodes (each with 8 GPUs): <syntaxhighlight lang="text"> compute-permanent-node-68 compute-permanent-node-285 compute-permanent-node-493 compute-permanent-node-625 compute-permanent-node-626 compute-permanent-node-749 compute-permanent-node-801 compute-permanent-node-580 </syntaxhighlight> When you ssh-in, you log in to the bastion node pure-caribou-bastion from which you can log in to any other node where you can test your code. == Reserving a GPU == Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run `gpunode`. You can see partition options by running `sinfo`. You might get an error like this: `groups: cannot find name for group ID 1506` But things should still run fine. Check with `nvidia-smi` [[Category:K-Scale]] 554106f928ebdcd704a5bcb31c89d8b23f6dcced 865 853 2024-05-06T02:58:37Z Ben 2 /* Reserving a GPU */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Cluster 1 === === Cluster 2 === The cluster has 8 available nodes (each with 8 GPUs): <syntaxhighlight lang="text"> compute-permanent-node-68 compute-permanent-node-285 compute-permanent-node-493 compute-permanent-node-625 compute-permanent-node-626 compute-permanent-node-749 compute-permanent-node-801 compute-permanent-node-580 </syntaxhighlight> When you ssh-in, you log in to the bastion node pure-caribou-bastion from which you can log in to any other node where you can test your code. == Reserving a GPU == Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. [[Category:K-Scale]] 1551cf14eb37ec1b4afa9dc2e329db2dd4704bcc 2 205 867 2024-05-06T03:39:00Z Dymaxion 22 Created page with "Kenji Sakaie is an incoming K-Scale intern and mechanical engineering student. {{infobox user | Name = Kenji Sakaie | Organization = K-Scale Labs and Franklin W. Olin College..." wikitext text/x-wiki Kenji Sakaie is an incoming K-Scale intern and mechanical engineering student. {{infobox user | Name = Kenji Sakaie | Organization = K-Scale Labs and Franklin W. Olin College of Engineering | Title = Intern }} 31edad4910e4da28e3584ef4e6eba07b36f67ed7 869 867 2024-05-06T03:41:18Z Dymaxion 22 wikitext text/x-wiki Kenji Sakaie is an incoming K-Scale intern and mechanical engineering student. {{infobox user }} 2518588e154c7b12509c61807ed63151664f126c 870 869 2024-05-06T03:44:29Z Dymaxion 22 wikitext text/x-wiki Kenji Sakaie is an incoming K-Scale intern and mechanical engineering student. {{infobox person | name = Kenji Sakaie | organization = K-Scale Labs | title = Technical Intern }} [[Category: K-Scale Employees]] 405768c9cfb9d33bd79d7aa27fa89f4386921534 10 206 868 2024-05-06T03:39:25Z Dymaxion 22 Created page with "{{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }}" wikitext text/x-wiki {{infobox company | name = Name | country = Country | website_link = https://link.com/ | robots = [[Stompy]] }} 788ee7706cdb15f6d9fd55449a0959a0d1ceff65 0 207 872 2024-05-06T04:09:57Z Dymaxion 22 Created page with "Robonaut2 was the first humanoid robot to be launched into space. It was developed by the NASA Johnson Space Center and launched in 2011, tested about the International Space..." wikitext text/x-wiki Robonaut2 was the first humanoid robot to be launched into space. It was developed by the NASA Johnson Space Center and launched in 2011, tested about the International Space Station, and returned to Earth for study in 2018. Robonaut2 was built to test applications of humanoid robots to augment crewed space missions. {{infobox robot | name = Robonaut2 | organization = [[NASA]] | video_link = https://www.youtube.com/watch?v=mtl48NOtyg0 | cost = $2,500,000 | height = 3' 4" | weight = 330 lbs | speed = 7 feet per second | lift_force = 40 lbs | number_made = 1 | status = Decommissioned }} [[Category:Robots]] 9c361afa851b4daa47e5d9165a1818ae47bcf046 0 170 874 715 2024-05-06T04:49:52Z Ben 2 wikitext text/x-wiki '''ODrive''' is a precision motor controller that is highly applicable in robotics domains, CNC machines, and other areas demanding high-performance motor control. It provides accurate control over hobby-grade and industrial electric motors. {{infobox actuator | name = ODrive Motor Controller | manufacturer = ODrive Robotics | cost = | purchase_link = https://odriverobotics.com/ | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = | gear_type = }} == Product Description == ODrive is primarily used for the control of electric motors with precision. The device is well-suited for accurate control of both hobby-grade and industrial motors. The controller has been specially designed for applications demanding high performance. The flexibility of the ODrive controller allows it to be used in a variety of applications including, but not limited to, robotics and CNC machines. == Industrial Application == The ODrive motor controller finds its application in numerous sectors, particularly in robotics and CNC machines. This is mainly due to its flexibility, which facilitates control over a variety of motors ranging from hobby-grade to industrial ones. Furthermore, the precision and high performance offered by the controller make it a preferred choice for professionals. == Technical Specifications == While the exact technical specifications of the ODrive motor controller can vary depending on the model, the controller is generally characterized by the following features: * High performance motor control: Ensures precision and accuracy. * Flexible interface: Facilitates easy integration with a variety of motors. * Compatible with various types of electric motors: Ensures wide application. == References == <references /> [[Category:Actuators]] 3ac8e5b196611a7d54a83f47a020a5998001227f 0 13 883 63 2024-05-06T17:59:10Z Vrtnis 21 wikitext text/x-wiki This is a course taught by Russ Tedrake at MIT. https://underactuated.csail.mit.edu/ [[:Category:Courses]] 60768f68f812a6c589c274fc03d09dd3780f2c67 885 883 2024-05-06T18:01:11Z Vrtnis 21 wikitext text/x-wiki This is a course taught by Russ Tedrake at MIT. https://underactuated.csail.mit.edu/ [[Category:Courses]] e39eec07ba37c96133c098a891395dbe79290bed 14 208 884 2024-05-06T17:59:37Z Vrtnis 21 Created blank page wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 1 886 875 2024-05-06T18:02:48Z Vrtnis 21 /* Resources */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |} 8960969842789dc22dfaaef248914a05133390f0 887 886 2024-05-06T20:33:27Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |} 61d0053b0a71c847d2d1b4acd0a86212ec2e55b8 922 887 2024-05-07T02:33:45Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |} 8db2633bdd9bf2ea653caaeca47a8d6f08ea2c9e 927 922 2024-05-07T04:49:10Z 185.169.0.83 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |} 9cf89c7b075cd779c7bc32ce60981f7fef320b71 939 927 2024-05-07T19:55:24Z Vrtnis 21 /* Resources */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |} 87d2e28f0ed4c1b0f2ce2062d8339112a5159bc3 942 939 2024-05-07T20:04:33Z Vrtnis 21 /* List of Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | Precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |} e02cd5e623f5a0cfec8ab70a9581a9f6bfe3f62e 943 942 2024-05-07T20:05:02Z Vrtnis 21 /* List of Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |} bf013f060fbe89a9f1fa7d35775ab7204417e5ff 0 209 888 2024-05-06T20:49:22Z Ben 2 Created page with "This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM * '''Power supply''' ** Assume we will have 48 volt power to the bo..." wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] 3b469fce1ea2dbec0c0effd712b227a99751b5b1 889 888 2024-05-06T21:03:38Z 32.213.80.127 0 /* Requirements */ add mcu's wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM *** Other options: **** Infineon XMC4800 **** NXP LPC1549 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] 3b8d0076d3fb5beb7d4ad4947c73dfc22cd68490 890 889 2024-05-06T21:05:08Z Ben 2 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Infineon XMC4800 *** NXP LPC1549 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] d20f56b85fabeb5447edc5290f4885b4834cc736 891 890 2024-05-06T21:06:05Z Ben 2 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> *** NXP LPC1549 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 5ec008f43be10466600b34335870ecee30585804 892 891 2024-05-06T21:07:50Z Ben 2 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial- microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> *** NXP LPC1549 *** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 1f7c438cf834a226e3bfdf55ae627dacbf36efd5 893 892 2024-05-06T21:11:58Z Ben 2 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> *** NXP LPC1549 *** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> c3fc8681c5d61d45912e60fd6583d6b3a81890e8 894 893 2024-05-06T21:13:23Z Matt 16 /* Requirements */ add sub categories wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549 *** MCU's **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 31d6aa47246c99af94a15b100c909c026c405068 895 894 2024-05-06T21:17:49Z Matt 16 /* Requirements */ add ref for NXP wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCU's **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 2f2e5a74c8fcc84da4c70ab174eeead8120aa137 896 895 2024-05-06T21:18:53Z Matt 16 /* Requirements */ add CAN transceiver wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCU's **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> a85d76abe9d7eee86c4c10889d1e9c5eede075e4 897 896 2024-05-06T21:23:01Z Matt 16 /* Requirements */ Add temp sensor wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCU's **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Relative position ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> f910a94cfb3cc27ec9fd3b2e96a5aaefb7bd8e3d 898 897 2024-05-06T21:25:29Z Ben 2 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCUs **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position ** Current sensing * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> e811e23502b08dcec5dd914faaffcde9efcecba1 899 898 2024-05-06T21:28:39Z Matt 16 /* Requirements */ Add current sensing IC + "Other" sub sections wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCUs **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 0a890c71d0371d5ed86548f2d8e57437571168db 900 899 2024-05-06T21:38:58Z Matt 16 /* Requirements */ add hell effect sensors for position wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCUs **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 2374c7b47d6d0ee0c9efbd2ac272e3d99d0a4d9f 901 900 2024-05-06T21:43:58Z Ben 2 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other options: *** Integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCUs **** RP2040 **** STM32F405RG ***** Used by [[ODrive]] * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> fe6fc2ef0f9ee10ae55ffc04eca15c44fc5d4b65 903 901 2024-05-06T21:45:20Z Matt 16 /* Requirements */ formatting wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** Integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> *** MCUs **** RP2040 **** STM32F405RG ***** Used by [[ODrive]] * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 086702a57181f686b016b1eebebb925176339fdf 904 903 2024-05-06T21:47:32Z Matt 16 /* Requirements */ fix odrive category placement wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** Integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG ***** Used by [[ODrive]] *** MCUs (without integrated motor control) **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 2d45a9c347a98d982b8e90a55716a975d1fd4596 905 904 2024-05-06T21:48:03Z Matt 16 /* Requirements */ wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** Integrated motor control MCUs **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG ***** Used by [[ODrive]] ***MCUs (without integrated motor control) **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 0727c86a1b7f5cfda296eae976395868014ed8d3 906 905 2024-05-06T21:54:17Z Matt 16 /* Requirements */ add comparison for mcus wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** '''Integrated motor control MCUs''' **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG ***** Used by [[ODrive]] *** For high-precision and complex control tasks with an emphasis on real-time networking and performance, the Infineon XMC4800 is suitable. *** If budget and simplicity are key considerations, and less intensive processing is required, the NXP LPC1549 offers a good balance. *** For a balance between performance and ecosystem support, with flexibility in hardware and software, the STM32F405RG is an excellent choice. *** '''MCUs (without integrated motor control)''' **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> d4052092f772f00e99ce3875cfe41f22f61bee2e 907 906 2024-05-06T22:08:37Z Matt 16 /* Requirements */ add ref for STG wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** '''Integrated motor control MCUs''' **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG<ref>https://www.lcsc.com/product-detail/Microcontroller-Units-MCUs-MPUs-SOCs_STMicroelectronics-STM32F405RGT6_C15742.html</ref> ***** Used by [[ODrive]] *** For high-precision and complex control tasks with an emphasis on real-time networking and performance, the Infineon XMC4800 is suitable. *** If budget and simplicity are key considerations, and less intensive processing is required, the NXP LPC1549 offers a good balance. *** For a balance between performance and ecosystem support, with flexibility in hardware and software, the STM32F405RG is an excellent choice. *** '''MCUs (without integrated motor control)''' **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> b2d8af72be49bbb49fd7131465050563668e6598 908 907 2024-05-06T22:23:58Z Matt 16 /* Requirements */ add xmc4500 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** '''Integrated motor control MCUs''' **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** Infineon XMC4500<ref>https://www.digikey.com/en/products/detail/infineon-technologies/XMC4500F100K1024ACXQSA1/4807912</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG<ref>https://www.lcsc.com/product-detail/Microcontroller-Units-MCUs-MPUs-SOCs_STMicroelectronics-STM32F405RGT6_C15742.html</ref> ***** Used by [[ODrive]] *** For high-precision and complex control tasks with an emphasis on real-time networking and performance, the Infineon XMC4800 is suitable. *** If budget and simplicity are key considerations, and less intensive processing is required, the NXP LPC1549 offers a good balance. *** For a balance between performance and ecosystem support, with flexibility in hardware and software, the STM32F405RG is an excellent choice. *** '''MCUs (without integrated motor control)''' **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> 587e53137f3ade83b366ff4a0ce227ef8c0b7f59 909 908 2024-05-06T22:27:27Z Matt 16 /* Requirements */ Update descriptions for when to use certain MCU's wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** '''Integrated motor control MCUs''' **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** Infineon XMC4500<ref>https://www.digikey.com/en/products/detail/infineon-technologies/XMC4500F100K1024ACXQSA1/4807912</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG<ref>https://www.lcsc.com/product-detail/Microcontroller-Units-MCUs-MPUs-SOCs_STMicroelectronics-STM32F405RGT6_C15742.html</ref> ***** Used by [[ODrive]] *** For high-precision and complex control tasks with an emphasis on real-time networking and performance, the Infineon XMC4800 is suitable, especially if you need robust communication features like EtherCAT for interconnected device control. *** If budget and simplicity are key considerations, and less intensive processing is required, the NXP LPC1549 offers a good balance. It provides essential capabilities for motor control without the complexities of advanced networking, making it ideal for straightforward applications. *** For a balance between performance and ecosystem support, with flexibility in hardware and software, the STM32F405RG is an excellent choice. It offers a powerful ARM Cortex-M4 processor and a rich development ecosystem, suitable for developers looking for extensive community support and flexibility. *** When high-end communication like EtherCAT isn't a necessity but robust control capabilities are still required, the Infineon XMC4500 is a strong contender. It provides similar processing power and peripheral support as the XMC4800 but without the integrated networking capabilities, which can be advantageous for applications focusing solely on control tasks without the need for complex network communications. *** '''MCUs (without integrated motor control)''' **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> c92bec144debd491657ede6657df21ce5fa419d4 916 909 2024-05-06T23:54:58Z Ben 2 wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** '''Integrated motor control MCUs''' **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** Infineon XMC4500<ref>https://www.digikey.com/en/products/detail/infineon-technologies/XMC4500F100K1024ACXQSA1/4807912</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG<ref>https://www.lcsc.com/product-detail/Microcontroller-Units-MCUs-MPUs-SOCs_STMicroelectronics-STM32F405RGT6_C15742.html</ref> ***** Used by [[ODrive]] *** For high-precision and complex control tasks with an emphasis on real-time networking and performance, the Infineon XMC4800 is suitable, especially if you need robust communication features like EtherCAT for interconnected device control. *** If budget and simplicity are key considerations, and less intensive processing is required, the NXP LPC1549 offers a good balance. It provides essential capabilities for motor control without the complexities of advanced networking, making it ideal for straightforward applications. *** For a balance between performance and ecosystem support, with flexibility in hardware and software, the STM32F405RG is an excellent choice. It offers a powerful ARM Cortex-M4 processor and a rich development ecosystem, suitable for developers looking for extensive community support and flexibility. *** When high-end communication like EtherCAT isn't a necessity but robust control capabilities are still required, the Infineon XMC4500 is a strong contender. It provides similar processing power and peripheral support as the XMC4800 but without the integrated networking capabilities, which can be advantageous for applications focusing solely on control tasks without the need for complex network communications. *** '''MCUs (without integrated motor control)''' **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 Mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> ff61892576dfffa43c0ad22d821c2537c090263d 947 916 2024-05-08T14:53:51Z Matt 16 /* Design */ wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** '''Integrated motor control MCUs''' **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** Infineon XMC4500<ref>https://www.digikey.com/en/products/detail/infineon-technologies/XMC4500F100K1024ACXQSA1/4807912</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG<ref>https://www.lcsc.com/product-detail/Microcontroller-Units-MCUs-MPUs-SOCs_STMicroelectronics-STM32F405RGT6_C15742.html</ref> ***** Used by [[ODrive]] *** For high-precision and complex control tasks with an emphasis on real-time networking and performance, the Infineon XMC4800 is suitable, especially if you need robust communication features like EtherCAT for interconnected device control. *** If budget and simplicity are key considerations, and less intensive processing is required, the NXP LPC1549 offers a good balance. It provides essential capabilities for motor control without the complexities of advanced networking, making it ideal for straightforward applications. *** For a balance between performance and ecosystem support, with flexibility in hardware and software, the STM32F405RG is an excellent choice. It offers a powerful ARM Cortex-M4 processor and a rich development ecosystem, suitable for developers looking for extensive community support and flexibility. *** When high-end communication like EtherCAT isn't a necessity but robust control capabilities are still required, the Infineon XMC4500 is a strong contender. It provides similar processing power and peripheral support as the XMC4800 but without the integrated networking capabilities, which can be advantageous for applications focusing solely on control tasks without the need for complex network communications. *** '''MCUs (without integrated motor control)''' **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 Mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] *[ Flux link here ] * Component Pins ** LCD ** CAN Transceiver ** Push Buttons === References === <references /> 97835758ff5fe62d265898bda122261c2c6d0ef9 948 947 2024-05-08T15:02:16Z Matt 16 /* Design */ wikitext text/x-wiki This is the K-Scale Motor Controller design document. === Requirements === * '''Microcontroller''' ** STM ** Other: *** '''Integrated motor control MCUs''' **** Infineon XMC4800<ref>https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/</ref> **** Infineon XMC4500<ref>https://www.digikey.com/en/products/detail/infineon-technologies/XMC4500F100K1024ACXQSA1/4807912</ref> **** NXP LPC1549<ref>https://www.digikey.com/en/products/detail/nxp-usa-inc/LPC1549JBD64QL/4696352?utm_adgroup=&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Medium%20ROAS%20Categories&utm_term=&utm_content=&utm_id=go_cmp-20223376311_adg-_ad-__dev-c_ext-_prd-_sig-Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB&gad_source=1&gclid=Cj0KCQjw_-GxBhC1ARIsADGgDjuYk7Hf3F-L_hoQ_4E-fPfjELknu3EAmm9IstEnc92wSAbIMEt0UZAaAsneEALw_wcB</ref> **** STM32F405RG<ref>https://www.lcsc.com/product-detail/Microcontroller-Units-MCUs-MPUs-SOCs_STMicroelectronics-STM32F405RGT6_C15742.html</ref> ***** Used by [[ODrive]] *** For high-precision and complex control tasks with an emphasis on real-time networking and performance, the Infineon XMC4800 is suitable, especially if you need robust communication features like EtherCAT for interconnected device control. *** If budget and simplicity are key considerations, and less intensive processing is required, the NXP LPC1549 offers a good balance. It provides essential capabilities for motor control without the complexities of advanced networking, making it ideal for straightforward applications. *** For a balance between performance and ecosystem support, with flexibility in hardware and software, the STM32F405RG is an excellent choice. It offers a powerful ARM Cortex-M4 processor and a rich development ecosystem, suitable for developers looking for extensive community support and flexibility. *** When high-end communication like EtherCAT isn't a necessity but robust control capabilities are still required, the Infineon XMC4500 is a strong contender. It provides similar processing power and peripheral support as the XMC4800 but without the integrated networking capabilities, which can be advantageous for applications focusing solely on control tasks without the need for complex network communications. *** '''MCUs (without integrated motor control)''' **** RP2040 * '''Power supply''' ** Assume we will have 48 volt power to the board * '''Communication''' ** CAN bus *** Other **** Texas Instruments ISO1050DUBR <ref>https://www.lcsc.com/product-detail/Isolated-CAN-Transceivers_Texas-Instruments-ISO1050DUBR_C16428.html</ref> ** 1 Mbps * '''Sensing''' ** Temperature *** Other **** Maxim Integrated DS18B20 <ref>https://www.lcsc.com/product-detail/Temperature-Sensors_Maxim-Integrated-DS18B20-T-R_C880672.html</ref> ** Absolute (single turn) position *** Other **** Infineon TLE5012B <ref> https://www.lcsc.com/product-detail/Position-Sensor_Infineon-Technologies-TLE5012B-E3005_C539928.html</ref> **** AMS AS5047P <ref> https://www.lcsc.com/product-detail/Position-Sensor_AMS-AS5047P-ATSM_C962063.html </ref> ** Current sensing *** Other **** Allegro ACS770 <ref> https://www.lcsc.com/product-detail/Current-Sensors_Allegro-MicroSystems-LLC-ACS770LCB-050U-PFF-T_C696104.html </ref> * '''Programming port''' ** USB C === Design === * [https://github.com/kscalelabs/motor-controller See this repository] === References === <references /> ff61892576dfffa43c0ad22d821c2537c090263d 0 170 902 874 2024-05-06T21:44:33Z Ben 2 wikitext text/x-wiki '''ODrive''' is a precision motor controller that is highly applicable in robotics domains, CNC machines, and other areas demanding high-performance motor control. It provides accurate control over hobby-grade and industrial electric motors. {{infobox actuator | name = ODrive Motor Controller | manufacturer = ODrive Robotics | cost = | purchase_link = https://odriverobotics.com/ | nominal_torque = | peak_torque = | weight = | dimensions = | gear_ratio = | voltage = | cad_link = | interface = | gear_type = }} == Product Description == ODrive is primarily used for the control of electric motors with precision. The device is well-suited for accurate control of both hobby-grade and industrial motors. The controller has been specially designed for applications demanding high performance. The flexibility of the ODrive controller allows it to be used in a variety of applications including, but not limited to, robotics and CNC machines. == Industrial Application == The ODrive motor controller finds its application in numerous sectors, particularly in robotics and CNC machines. This is mainly due to its flexibility, which facilitates control over a variety of motors ranging from hobby-grade to industrial ones. Furthermore, the precision and high performance offered by the controller make it a preferred choice for professionals. == Technical Specifications == While the exact technical specifications of the ODrive motor controller can vary depending on the model, the controller is generally characterized by the following features: * High performance motor control: Ensures precision and accuracy. * Flexible interface: Facilitates easy integration with a variety of motors. * Compatible with various types of electric motors: Ensures wide application. == Integrated Circuits == * '''MCU''': STM32F405RG<ref>https://www.st.com/en/microcontrollers-microprocessors/stm32f405rg.html</ref> == References == <references /> [[Category:Actuators]] aa149883cde57b753c3591906cd15b7dc9f3e64e 6 210 910 2024-05-06T22:29:53Z Ben 2 wikitext text/x-wiki RMD X8 PCB bottom view 5a8a86c86a7c660035ccdf6c7d0a09bdea993583 6 211 911 2024-05-06T22:30:34Z Ben 2 wikitext text/x-wiki RMD X6 exploded view 38269032403fbebefab2d559789b39a394a3720f 6 212 914 2024-05-06T22:32:51Z Ben 2 wikitext text/x-wiki RMD X-Series CAN Bus 6d29aadd6ac32c09f614c89058d4601f8c60ef80 0 213 923 2024-05-07T02:37:43Z Modeless 7 Created page with "LASER Robotics seems to be founded by people from USC's Dynamic Robotics and Control Laboratory. They have a humanoid called HECTOR v2.<ref>https://laser-robotics.com/hector-v..." wikitext text/x-wiki LASER Robotics seems to be founded by people from USC's Dynamic Robotics and Control Laboratory. They have a humanoid called HECTOR v2.<ref>https://laser-robotics.com/hector-v2/</ref> 2626f36cafb6480c2f97b92c3a28b08f1a173d84 924 923 2024-05-07T02:41:03Z Modeless 7 wikitext text/x-wiki LASER Robotics seems to be founded by people from USC's Dynamic Robotics and Control Laboratory. They have a small .85m tall humanoid called [[HECTOR V2]] with open source control and simulation software aimed at research.<ref>https://laser-robotics.com/hector-v2/</ref> 64b5062d242f111e356277e0464145b1a7f0e52d 933 924 2024-05-07T05:12:18Z Ben 2 wikitext text/x-wiki LASER Robotics seems to be founded by people from USC's Dynamic Robotics and Control Laboratory. They have a small .85m tall humanoid called [[HECTOR V2]] with open source control and simulation software aimed at research.<ref>https://laser-robotics.com/hector-v2/</ref> === References === <references/> e803bac2792a515757969bc5d70c6b1735aaa900 0 214 928 2024-05-07T04:49:43Z 185.169.0.83 0 Created page with "A low-cost backdrivable 7-DOF robotic manipulator.<ref>https://berkeleyopenarms.github.io/</ref> === References === <references/>" wikitext text/x-wiki A low-cost backdrivable 7-DOF robotic manipulator.<ref>https://berkeleyopenarms.github.io/</ref> === References === <references/> 2558b9f89a4939aca41dffb330ee4003e36b38ab 0 16 930 865 2024-05-07T05:08:49Z Ben 2 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Cluster 1 === === Cluster 2 === The cluster has 8 available nodes (each with 8 GPUs): <syntaxhighlight lang="text"> compute-permanent-node-68 compute-permanent-node-285 compute-permanent-node-493 compute-permanent-node-625 compute-permanent-node-626 compute-permanent-node-749 compute-permanent-node-801 compute-permanent-node-580 </syntaxhighlight> When you ssh-in, you log in to the bastion node pure-caribou-bastion from which you can log in to any other node where you can test your code. == Reserving a GPU == Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. === Useful Commands === Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] 1fd3b99c33a47197d00f66fa784a036813390556 0 215 931 2024-05-07T05:11:31Z Ben 2 Created page with "Open-source humanoid robot from USC.<ref>https://github.com/DRCL-USC/Hector_Simulation</ref> === References === <references/>" wikitext text/x-wiki Open-source humanoid robot from USC.<ref>https://github.com/DRCL-USC/Hector_Simulation</ref> === References === <references/> 1a3dbfa9a8cedc111c3a882f8ee1009b58969c83 932 931 2024-05-07T05:11:48Z Ben 2 wikitext text/x-wiki Open-source humanoid robot from USC.<ref>https://github.com/DRCL-USC/Hector_Simulation</ref><ref>https://laser-robotics.com/hector-v2/</ref> === References === <references/> 7977068d4016c1acee3e50cb9050c71f3410584d 0 193 940 882 2024-05-07T19:57:01Z Vrtnis 21 /* Experimenting with Your Humanoid Robot */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the [https://humanoids.wiki/w/MIT_Cheetah MIT Cheetah] actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost.If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. More resources are available at [https://humanoids.wiki/w/Learning_algorithms] == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to an open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs. Check out [https://humanoids.wiki/w/Stompy Stompy!] === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. d2813031631c0d09b1c6e27bfddb1632e70152fa 941 940 2024-05-07T19:58:15Z Vrtnis 21 /* Best Practices for Virtual Testing */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the [https://humanoids.wiki/w/MIT_Cheetah MIT Cheetah] actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost.If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== Reinforcement Learning Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. More resources are available at [https://humanoids.wiki/w/Learning_algorithms Learning Algorithms] == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to an open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs. Check out [https://humanoids.wiki/w/Stompy Stompy!] === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 4e144263418dd20bdd1d540f94b223463b8597be 0 216 946 2024-05-08T01:37:08Z Ben 2 Created page with "Tracking document for [[User:Ben]]'s random ideas for things to work on. * Discord bot to automatically post robotics and relevant machine learning papers to the K-Scale Labs..." wikitext text/x-wiki Tracking document for [[User:Ben]]'s random ideas for things to work on. * Discord bot to automatically post robotics and relevant machine learning papers to the K-Scale Labs Discord group ** Maybe summarize the paper results using ChatGPT? cf833af97d410caa2ed68f311594c1f135100868 0 217 949 2024-05-08T15:07:11Z Matt 16 Add motor programmer page wikitext text/x-wiki === Purpose === aka "The Doodad" === Design === [https://www.flux.ai/k-scale-labs/motor-programmer?editor=schematic&embed=1 Flux Design Link] * Component Pins ** LCD ** CAN Transceiver ** Push Buttons === Components === * MCU * LCD ** HD44780 d4bd2bf48530ee4145a4c86450d53e4064b8d7a8 950 949 2024-05-08T15:09:49Z Matt 16 /* Components */ wikitext text/x-wiki === Purpose === aka "The Doodad" === Design === [https://www.flux.ai/k-scale-labs/motor-programmer?editor=schematic&embed=1 Flux Design Link] * Component Pins ** LCD ** CAN Transceiver ** Push Buttons === Components === * MCU * LCD ** Hitachi HD44780 658142f7995123d5a3e03c4d5ebec4eea521ae7d 0 217 951 950 2024-05-08T15:30:36Z Matt 16 /* Design */ wikitext text/x-wiki === Purpose === aka "The Doodad" === Design === [https://www.flux.ai/k-scale-labs/motor-programmer?editor=schematic&embed=1 Flux Design Link] * Component Pins ** LCD ** CAN Transceiver ** Push Buttons ** STM *** Internal Pull up resistor required === Components === * MCU * LCD ** Hitachi HD44780 f565fba130c4d6be3d79c578cf982f242eecfe67 952 951 2024-05-08T15:34:47Z Matt 16 /* Design */ wikitext text/x-wiki === Purpose === aka "The Doodad" === Design === [https://www.flux.ai/k-scale-labs/motor-programmer?editor=schematic&embed=1 Flux Design Link] * Component Pins ** LCD ** CAN Transceiver ** Push Buttons ** STM *** Internal pull down resistor required === Components === * MCU * LCD ** Hitachi HD44780 3479fb6c46775ba28425e5a578d12a81644b4b41 0 218 956 2024-05-08T21:12:57Z Tom 23 Created page with "Notes on programming/interfacing with Jetson Orin hardware." wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. 4402ff9a963e565cb37c8994f02ffe654f7a70e3 957 956 2024-05-08T21:21:36Z Tom 23 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. # Upgrading AGX to Jetson Linux 36.3 ## BSP approach (avoids SDK Manager) * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from https://developer.nvidia.com/embedded/jetson-linux * Unpack (as root, get used to doing most of this as root), preserving privileges - `tar xjpf ...` * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run './apply_binaries.sh' from the BSP - Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the `lsusb` * Run './nvautoflash.sh' * Watch for few minutes, typically it crashes early, then go for lunch. b8cfcabc797bc60e0fdf4b7ec9eea17488ff9101 960 957 2024-05-08T21:47:45Z Budzianowski 19 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. # Upgrading AGX to Jetson Linux 36.3 ## BSP approach (avoids SDK Manager) * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from https://developer.nvidia.com/embedded/jetson-linux * Unpack (as root, get used to doing most of this as root), preserving privileges - `tar xjpf ...` * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run './apply_binaries.sh' from the BSP - Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the `lsusb` * Run './nvautoflash.sh' * Watch for few minutes, typically it crashes early, then go for lunch. [[Firmware]] 0d0d7525fce2f6e905beb82228b66a75b7df005c 961 960 2024-05-08T21:48:41Z Budzianowski 19 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. # Upgrading AGX to Jetson Linux 36.3 ## BSP approach (avoids SDK Manager) * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from https://developer.nvidia.com/embedded/jetson-linux * Unpack (as root, get used to doing most of this as root), preserving privileges - `tar xjpf ...` * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run './apply_binaries.sh' from the BSP - Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the `lsusb` * Run './nvautoflash.sh' * Watch for few minutes, typically it crashes early, then go for lunch. [[Category: Firmware]] aa3877daafd814bd7b737663faba30147c33f9a4 965 961 2024-05-09T04:29:23Z Ben 2 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> * Run <code>./nvautoflash.sh</code> * Watch for few minutes, typically it crashes early, then go for lunch. [[Category: Firmware]] 65a28872b9149b997b827924eb52f67d1c44a6be 982 965 2024-05-10T22:23:22Z Budzianowski 19 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run sudo ./tools/l4t_flash_prerequisites.sh * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> under NVIDIA CORP. APX * Run <code>./flash.sh </code> Different options for different usecases(https://docs.nvidia.com/jetson/archives/r36.3/DeveloperGuide/IN/QuickStart.html#in-quickstart) Jetson AGX Orin Developer Kit (eMMC): $ sudo ./flash.sh jetson-agx-orin-devkit internal * Watch for few minutes, typically it crashes early, then go for lunch. [[Category: Firmware]] 071f59c04c2524b3905f281963f6223a0d3ce3a2 2 219 958 2024-05-08T21:23:10Z Tom 23 Created page with " - Tom "qdot.me" Mloduchowski" wikitext text/x-wiki - Tom "qdot.me" Mloduchowski fe02db12102ee6a58d648f1d48ce795280313f5b 0 1 959 943 2024-05-08T21:42:54Z 108.211.178.220 0 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |} 8bcfb12984f64974d144a143e4140b3fc1108963 963 959 2024-05-09T03:28:36Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |} 18aade77769ec45ea72eb74dbfe7a78b354ffb9e 996 963 2024-05-13T07:26:40Z Ben 2 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |} 6ccb0a7f9b51c2682bc6937875c6ce219294038e 14 220 962 2024-05-08T21:48:54Z Budzianowski 19 Created blank page wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 221 964 2024-05-09T03:29:43Z Modeless 7 Created page with "German company NEURA Robotics announced a full humanoid called 4NE-1.<ref>https://neura-robotics.com/products/4ne-1</ref><ref>https://www.youtube.com/watch?v=tUgV4RylVU0</ref>" wikitext text/x-wiki German company NEURA Robotics announced a full humanoid called 4NE-1.<ref>https://neura-robotics.com/products/4ne-1</ref><ref>https://www.youtube.com/watch?v=tUgV4RylVU0</ref> f1eb449e58778ae2c6d0288e10d2833610669873 2 222 966 2024-05-09T10:46:06Z Vedant 24 Created page with "{| class="wikitable" |- ! Vedant Jhawar !! |- | Name || Vedant Jhawar |- | Organization || K-Scale Labs |- | Title || Technical Intern |- |}" wikitext text/x-wiki {| class="wikitable" |- ! Vedant Jhawar !! |- | Name || Vedant Jhawar |- | Organization || K-Scale Labs |- | Title || Technical Intern |- |} 276b737332680efefef379118a77a1e1318a416a 967 966 2024-05-09T10:47:11Z Vedant 24 wikitext text/x-wiki {{infobox person | name = Vedant Jhawar | organization = [[K-Scale Labs]] | title = Technical Intern }} [[Category: K-Scale Employees]] 0c2b35a23e45810856abe7bfb28bb55d3093726b 0 20 968 225 2024-05-09T16:05:24Z Budzianowski 19 Budzianowski moved page [[Jetson Orin]] to [[CAN/IMU/Cameras with Jetson Orin]] wikitext text/x-wiki The Jetson Orin is a development board from Nvidia. === CAN Bus === See [https://docs.nvidia.com/jetson/archives/r34.1/DeveloperGuide/text/HR/ControllerAreaNetworkCan.html here] for notes on configuring the CAN bus for the Jetson. [[File:Can bus connections 2.png|none|200px|thumb]] Install dependencies: <syntaxhighlight lang="bash"> sudo add-apt-repository ppa:deadsnakes/ppa sudo apt update sudo apt upgrade sudo apt install g++ python3.11-dev </syntaxhighlight> Initialize the CAN bus on startup: <syntaxhighlight lang="bash"> #!/bin/bash # Set pinmux. busybox devmem 0x0c303000 32 0x0000C400 busybox devmem 0x0c303008 32 0x0000C458 busybox devmem 0x0c303010 32 0x0000C400 busybox devmem 0x0c303018 32 0x0000C458 # Install modules. modprobe can modprobe can_raw modprobe mttcan # Turn off CAN. ip link set down can0 ip link set down can1 # Set parameters. ip link set can0 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off ip link set can1 type can bitrate 1000000 dbitrate 1000000 berr-reporting on fd on loopback off # Turn on CAN. ip link set up can0 ip link set up can1 </syntaxhighlight> You can run this script automatically on startup by writing a service configuration to (for example) <code>/etc/systemd/system/can_setup.service</code> <syntaxhighlight lang="text"> [Unit] Description=Initialize CAN Interfaces After=network.target [Service] Type=oneshot ExecStart=/opt/kscale/enable_can.sh RemainAfterExit=true [Install] WantedBy=multi-user.target </syntaxhighlight> To enable this, run: <syntaxhighlight lang="bash"> sudo systemctl enable can_setup sudo systemctl start can_setup </syntaxhighlight> === Cameras === ==== Arducam IMX 219 ==== * [https://www.arducam.com/product/arducam-imx219-multi-camera-kit-for-the-nvidia-jetson-agx-orin/ Product Page] ** Shipping was pretty fast ** Order a couple backup cameras because a couple of the cameras that they shipped came busted * [https://docs.arducam.com/Nvidia-Jetson-Camera/Nvidia-Jetson-Orin-Series/NVIDIA-Jetson-AGX-Orin/Quick-Start-Guide/ Quick start guide] Run the installation script: <syntaxhighlight lang="bash"> wget https://github.com/ArduCAM/MIPI_Camera/releases/download/v0.0.3/install_full.sh chmod u+x install_full.sh ./install_full.sh -m imx219 </syntaxhighlight> Supported kernel versions (see releases [https://github.com/ArduCAM/MIPI_Camera/releases here]): * <code>5.10.104-tegra-35.3.1</code> * <code>5.10.120-tegra-35.4.1</code> Install an older kernel from [https://developer.nvidia.com/embedded/jetson-linux-archive here]. This required downgrading to Ubuntu 20.04 (only changing <code>/etc/os-version</code>). Install dependencies: <syntaxhighlight lang="bash"> sudo apt update sudo apt install \ gstreamer1.0-tools \ gstreamer1.0-alsa \ gstreamer1.0-plugins-base \ gstreamer1.0-plugins-good \ gstreamer1.0-plugins-bad \ gstreamer1.0-plugins-ugly \ gstreamer1.0-libav sudo apt install libgstreamer1.0-dev \ libgstreamer-plugins-base1.0-dev \ libgstreamer-plugins-good1.0-dev \ libgstreamer-plugins-bad1.0-dev sudo apt install \ v4l-utils \ ffmpeg </syntaxhighlight> Make sure the camera shows up: <syntaxhighlight lang="bash"> v4l2-ctl --list-formats-ext </syntaxhighlight> Capture a frame from the camera: <syntaxhighlight lang="bash"> gst-launch-1.0 nvarguscamerasrc sensor-id=0 ! "video/x-raw(memory:NVMM), width=1280, height=720, framerate=60/1" ! nvvidconv ! jpegenc snapshot=TRUE ! filesink location=test.jpg </syntaxhighlight> Alternatively, use the following Python code: <syntaxhighlight lang="bash"> import cv2 gst_str = ( 'nvarguscamerasrc sensor-id=0 ! ' 'video/x-raw(memory:NVMM), width=(int)1280, height=(int)720, format=(string)NV12, framerate=(fraction)60/1 ! ' 'nvvidconv flip-method=0 ! ' 'video/x-raw, width=(int)1280, height=(int)720, format=(string)BGRx ! ' 'videoconvert ! ' 'video/x-raw, format=(string)BGR ! ' 'appsink' ) cap = cv2.VideoCapture(gst_str, cv2.CAP_GSTREAMER) while True: ret, frame = cap.read() if ret: print(frame) if cv2.waitKey(1) & 0xFF == ord('q'): break else: break cap.release() cv2.destroyAllWindows() </syntaxhighlight> === IMU === We're using the [https://ozzmaker.com/product/berryimu-accelerometer-gyroscope-magnetometer-barometricaltitude-sensor/ BerryIMU v3]. To use it, connect pin 3 on the Jetson to SDA and pin 5 to SCL for I2C bus 7. You can verify the connection is successful if the following command matches: <syntaxhighlight lang="bash"> $ sudo i2cdetect -y -r 7 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- 1c -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- 6a -- -- -- -- -- 70: -- -- -- -- -- -- -- 77 </syntaxhighlight> The equivalent command on the Raspberry Pi should use bus 1: <syntaxhighlight lang="bash"> sudo i2cdetect -y -r 1 </syntaxhighlight> The default addresses are: * <code>0x6A</code>: Gyroscope and accelerometer * <code>0x1C</code>: Magnetometer * <code>0x77</code>: Barometer [[Category: Hardware]] [[Category: Electronics]] 51030b57b98b338a704c3f6daae5082cb8b122c3 0 223 969 2024-05-09T16:05:26Z Budzianowski 19 Budzianowski moved page [[Jetson Orin]] to [[CAN/IMU/Cameras with Jetson Orin]] wikitext text/x-wiki #REDIRECT [[CAN/IMU/Cameras with Jetson Orin]] cc10907f7af0d9ca273063eeb79b951ec04e1607 0 224 970 2024-05-09T20:03:59Z Vrtnis 21 Created page with "== Isaac Sim Automator == The Isaac Sim Automator tool automates the deployment of Isaac Sim to public clouds, enabling efficient and scalable simulations. === Installation =..." wikitext text/x-wiki == Isaac Sim Automator == The Isaac Sim Automator tool automates the deployment of Isaac Sim to public clouds, enabling efficient and scalable simulations. === Installation === ==== Installing Docker ==== Ensure Docker is installed on your system for container management. For installation guidance, see [https://docs.docker.com/engine/install/ Docker Installation]. ==== Obtaining NGC API Key ==== Obtain an NGC API Key to download Docker images from NVIDIA NGC. This can be done at [https://ngc.nvidia.com/setup/api-key NGC API Key Setup]. ==== Building the Container ==== To build the Automator container, use the following command in your project root directory: <source lang="bash"> ./build </source> This builds and tags the Isaac Sim Automator container as 'isa'. === Usage === ==== Running Automator Commands ==== You can run the Automator commands in two ways: * Enter the automator container and run commands inside: <source lang="bash"> ./run ./somecommand </source> * Or, prepend the command with ./run: <source lang="bash"> ./run ./somecommand <parameters> </source> Examples include: <source lang="bash"> ./run ./deploy-aws ./run ./destroy my-deployment </source> ==== Deploying Isaac Sim ==== Choose the appropriate cloud provider (AWS, GCP, Azure, Alibaba Cloud) and follow the provided steps to deploy Isaac Sim using the automator container. == K-Scale Sim Library == The K-Scale Sim Library is built atop Isaac Gym for simulating Stompy, providing interfaces for defined tasks like walking and getting up. === Getting Started with K-Scale Sim Library === ==== Initial Setup ==== First, clone the K-Scale Sim Library repository and set up the environment: <source lang="bash"> git clone https://github.com/kscalelabs/sim.git cd sim conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </source> ==== Installing Dependencies ==== After setting up the base environment, download and install necessary third-party packages: <source lang="bash"> wget https://developer.nvidia.com/isaac-gym/IsaacGym_Preview_4_Package.tar.gz tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </source> === Running Experiments === ==== Setting Up Experiments ==== Download and prepare the Stompy model for experiments: <source lang="bash"> wget https://media.kscale.dev/stompy.tar.gz tar -xzvf stompy.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </source> ==== Training and Evaluation ==== * For leg-specific tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </source> * For full-body tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </source> Evaluate the models on CPU: <source lang="bash"> python sim/humanoid_gym/play.py --task=legs_ppo --sim_device=cpu </source> === Troubleshooting === Common issues and solutions for setting up and running Isaac Gym and K-Scale simulations. <source lang="bash"> git submodule update --init --recursive export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH sudo apt-get install vulkan1 </source> 9bed181ad0605992dde731edbad2cf05082e67fb 971 970 2024-05-09T20:10:38Z Vrtnis 21 /* Isaac Sim Automator */ wikitext text/x-wiki == Isaac Sim Automator == The '''Isaac Sim Automator''' is essentially your go-to tool for setting up and managing Isaac Sim on various cloud platforms, and it does all this with the help of Docker. Docker packages up an application with all its dependencies into a container that can run anywhere. This means you don't have to worry about your simulation behaving differently from one environment to another — whether it's on your laptop or in the cloud. Using the Automator is super straightforward. It comes with a set of command-line tools that make deploying to clouds like AWS, Google Cloud, Azure, and Alibaba Cloud almost as easy as sending an email. Here’s how you can deploy Isaac Sim to, say, AWS: <source lang="bash"> ./deploy-aws </source> Or if you're more into Google Cloud, you'd use: <source lang="bash"> ./deploy-gcp </source> These commands kick off the entire process to get Isaac Sim up and running on your preferred cloud service. They manage all the techie stuff like setting up resources and configuring instances, which means less hassle for you. In a nutshell, the Isaac Sim Automator is fantastic if you’re diving into the K-Scale Sim Library. It simplifies all the tricky cloud setup stuff, letting you focus on creating and running your simulations. Whether you're a researcher, a developer, or just someone playing around with simulations, this tool can save you a ton of time and effort. === Installation === ==== Installing Docker ==== Ensure Docker is installed on your system for container management. For installation guidance, see [https://docs.docker.com/engine/install/ Docker Installation]. ==== Obtaining NGC API Key ==== Obtain an NGC API Key to download Docker images from NVIDIA NGC. This can be done at [https://ngc.nvidia.com/setup/api-key NGC API Key Setup]. ==== Building the Container ==== To build the Automator container, use the following command in your project root directory: <source lang="bash"> ./build </source> This builds and tags the Isaac Sim Automator container as 'isa'. === Usage === ==== Running Automator Commands ==== You can run the Automator commands in two ways: * Enter the automator container and run commands inside: <source lang="bash"> ./run ./somecommand </source> * Or, prepend the command with ./run: <source lang="bash"> ./run ./somecommand <parameters> </source> Examples include: <source lang="bash"> ./run ./deploy-aws ./run ./destroy my-deployment </source> ==== Deploying Isaac Sim ==== Choose the appropriate cloud provider (AWS, GCP, Azure, Alibaba Cloud) and follow the provided steps to deploy Isaac Sim using the automator container. == K-Scale Sim Library == The K-Scale Sim Library is built atop Isaac Gym for simulating Stompy, providing interfaces for defined tasks like walking and getting up. === Getting Started with K-Scale Sim Library === ==== Initial Setup ==== First, clone the K-Scale Sim Library repository and set up the environment: <source lang="bash"> git clone https://github.com/kscalelabs/sim.git cd sim conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </source> ==== Installing Dependencies ==== After setting up the base environment, download and install necessary third-party packages: <source lang="bash"> wget https://developer.nvidia.com/isaac-gym/IsaacGym_Preview_4_Package.tar.gz tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </source> === Running Experiments === ==== Setting Up Experiments ==== Download and prepare the Stompy model for experiments: <source lang="bash"> wget https://media.kscale.dev/stompy.tar.gz tar -xzvf stompy.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </source> ==== Training and Evaluation ==== * For leg-specific tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </source> * For full-body tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </source> Evaluate the models on CPU: <source lang="bash"> python sim/humanoid_gym/play.py --task=legs_ppo --sim_device=cpu </source> === Troubleshooting === Common issues and solutions for setting up and running Isaac Gym and K-Scale simulations. <source lang="bash"> git submodule update --init --recursive export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH sudo apt-get install vulkan1 </source> a0917046510ae64286dcf1f69cba4042f6dde290 972 971 2024-05-09T20:24:22Z Vrtnis 21 wikitext text/x-wiki == Isaac Sim Automator == The '''Isaac Sim Automator''' is essentially your go-to tool for setting up and managing Isaac Sim on various cloud platforms, and it does all this with the help of Docker. Docker packages up an application with all its dependencies into a container that can run anywhere. This means you don't have to worry about your simulation behaving differently from one environment to another — whether it's on your laptop or in the cloud. Using the Automator is super straightforward. It comes with a set of command-line tools that make deploying to clouds like AWS, Google Cloud, Azure, and Alibaba Cloud. Here’s how you can deploy Isaac Sim to, say, AWS: <source lang="bash"> ./deploy-aws </source> Or if you're more into Google Cloud, you'd use: <source lang="bash"> ./deploy-gcp </source> These commands kick off the entire process to get Isaac Sim up and running on your preferred cloud service. They manage details like setting up resources and configuring instances. In a nutshell, the Isaac Sim Automator is fantastic if you’re diving into the K-Scale Sim Library. It simplifies cloud setup, letting you focus on creating and running your simulations. Whether you're a researcher, a developer, or a beginner to robotics, this tool can save you a ton of time and effort. === Installation === ==== Installing Docker ==== Ensure Docker is installed on your system for container management. For installation guidance, see [https://docs.docker.com/engine/install/ Docker Installation]. ==== Obtaining NGC API Key ==== Obtain an NGC API Key to download Docker images from NVIDIA NGC. This can be done at [https://ngc.nvidia.com/setup/api-key NGC API Key Setup]. ==== Building the Container ==== To build the Automator container, use the following command in your project root directory: <source lang="bash"> ./build </source> This builds and tags the Isaac Sim Automator container as 'isa'. === Usage === ==== Running Automator Commands ==== You can run the Automator commands in two ways: * Enter the automator container and run commands inside: <source lang="bash"> ./run ./somecommand </source> * Or, prepend the command with ./run: <source lang="bash"> ./run ./somecommand <parameters> </source> Examples include: <source lang="bash"> ./run ./deploy-aws ./run ./destroy my-deployment </source> ==== Deploying Isaac Sim ==== Choose the appropriate cloud provider (AWS, GCP, Azure, Alibaba Cloud) and follow the provided steps to deploy Isaac Sim using the automator container. == K-Scale Sim Library == The K-Scale Sim Library is built atop Isaac Gym for simulating Stompy, providing interfaces for defined tasks like walking and getting up. === Getting Started with K-Scale Sim Library === ==== Initial Setup ==== First, clone the K-Scale Sim Library repository and set up the environment: <source lang="bash"> git clone https://github.com/kscalelabs/sim.git cd sim conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </source> ==== Installing Dependencies ==== After setting up the base environment, download and install necessary third-party packages: <source lang="bash"> wget https://developer.nvidia.com/isaac-gym/IsaacGym_Preview_4_Package.tar.gz tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </source> === Running Experiments === ==== Setting Up Experiments ==== Download and prepare the Stompy model for experiments: <source lang="bash"> wget https://media.kscale.dev/stompy.tar.gz tar -xzvf stompy.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </source> ==== Training and Evaluation ==== * For leg-specific tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </source> * For full-body tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </source> Evaluate the models on CPU: <source lang="bash"> python sim/humanoid_gym/play.py --task=legs_ppo --sim_device=cpu </source> === Troubleshooting === Common issues and solutions for setting up and running Isaac Gym and K-Scale simulations. <source lang="bash"> git submodule update --init --recursive export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH sudo apt-get install vulkan1 </source> 7356970102cfdcd28f2ac67ecd6b5fd2127ab569 973 972 2024-05-09T20:51:56Z Vrtnis 21 /* Isaac Sim Automator */ wikitext text/x-wiki == Isaac Sim Automator == The [https://github.com/NVIDIA-Omniverse/IsaacSim-Automator/ Isaac Sim Automator] is essentially your go-to tool for setting up and managing Isaac Sim on various cloud platforms, and it does all this with the help of Docker. Docker packages up an application with all its dependencies into a container that can run anywhere. This means you don't have to worry about your simulation behaving differently from one environment to another — whether it's on your laptop or in the cloud. Using the Automator is super straightforward. It comes with a set of command-line tools that make deploying to clouds like AWS, Google Cloud, Azure, and Alibaba Cloud. Here’s how you can deploy Isaac Sim to, say, AWS: <source lang="bash"> ./deploy-aws </source> Or if you're more into Google Cloud, you'd use: <source lang="bash"> ./deploy-gcp </source> These commands kick off the entire process to get Isaac Sim up and running on your preferred cloud service. They manage details like setting up resources and configuring instances. In a nutshell, the Isaac Sim Automator is fantastic if you’re diving into the K-Scale Sim Library. It simplifies cloud setup, letting you focus on creating and running your simulations. Whether you're a researcher, a developer, or a beginner to robotics, this tool can save you a ton of time and effort. === Installation === ==== Installing Docker ==== Ensure Docker is installed on your system for container management. For installation guidance, see [https://docs.docker.com/engine/install/ Docker Installation]. ==== Obtaining NGC API Key ==== Obtain an NGC API Key to download Docker images from NVIDIA NGC. This can be done at [https://ngc.nvidia.com/setup/api-key NGC API Key Setup]. ==== Building the Container ==== To build the Automator container, use the following command in your project root directory: <source lang="bash"> ./build </source> This builds and tags the Isaac Sim Automator container as 'isa'. === Usage === ==== Running Automator Commands ==== You can run the Automator commands in two ways: * Enter the automator container and run commands inside: <source lang="bash"> ./run ./somecommand </source> * Or, prepend the command with ./run: <source lang="bash"> ./run ./somecommand <parameters> </source> Examples include: <source lang="bash"> ./run ./deploy-aws ./run ./destroy my-deployment </source> ==== Deploying Isaac Sim ==== Choose the appropriate cloud provider (AWS, GCP, Azure, Alibaba Cloud) and follow the provided steps to deploy Isaac Sim using the automator container. == K-Scale Sim Library == The K-Scale Sim Library is built atop Isaac Gym for simulating Stompy, providing interfaces for defined tasks like walking and getting up. === Getting Started with K-Scale Sim Library === ==== Initial Setup ==== First, clone the K-Scale Sim Library repository and set up the environment: <source lang="bash"> git clone https://github.com/kscalelabs/sim.git cd sim conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </source> ==== Installing Dependencies ==== After setting up the base environment, download and install necessary third-party packages: <source lang="bash"> wget https://developer.nvidia.com/isaac-gym/IsaacGym_Preview_4_Package.tar.gz tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </source> === Running Experiments === ==== Setting Up Experiments ==== Download and prepare the Stompy model for experiments: <source lang="bash"> wget https://media.kscale.dev/stompy.tar.gz tar -xzvf stompy.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </source> ==== Training and Evaluation ==== * For leg-specific tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </source> * For full-body tasks: <source lang="bash"> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </source> Evaluate the models on CPU: <source lang="bash"> python sim/humanoid_gym/play.py --task=legs_ppo --sim_device=cpu </source> === Troubleshooting === Common issues and solutions for setting up and running Isaac Gym and K-Scale simulations. <source lang="bash"> git submodule update --init --recursive export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH sudo apt-get install vulkan1 </source> 42f21ba77ab4c6c4f34c4d5f94aa12652ea662a1 2 225 974 2024-05-09T23:53:01Z Budzianowski 19 Created page with "hello" wikitext text/x-wiki hello aaf4c61ddcc5e8a2dabede0f3b482cd9aea9434d 975 974 2024-05-09T23:54:04Z Budzianowski 19 wikitext text/x-wiki [http://budzianowski.github.io/ Paweł] is one of the co-founders of [[K-Scale Labs]]. [[Category: K-Scale Employees]] 1594da6290ba47ad956555057bbdcbce62f26923 976 975 2024-05-09T23:54:41Z Budzianowski 19 wikitext text/x-wiki [http://budzianowski.github.io/ Paweł] is one of the co-founders of [[K-Scale Labs]]. AI, cycling and a tea with lemon. [[Category: K-Scale Employees]] 8dc76d8c7383ff260783b233032a787edde1af09 6 226 977 2024-05-10T00:29:25Z Ben 2 wikitext text/x-wiki Stompy 1 8bc1a9c3b2ff6d0f29975f88f8692c36f88a69f1 6 227 978 2024-05-10T00:29:40Z Ben 2 wikitext text/x-wiki Stompy 2 e0908a5eb185820dad543ca27d2df7f64a056f8e 6 228 979 2024-05-10T00:29:53Z Ben 2 wikitext text/x-wiki Stompy 3 156eb4ce99187583e5cb9c1e9eb6225d0ad0508e 6 229 980 2024-05-10T00:30:06Z Ben 2 wikitext text/x-wiki Stompy 4 67791452427b9b3b5d1bf0e3254fbc15ee179cb9 0 2 981 721 2024-05-10T00:30:39Z Ben 2 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. = Artwork = Here's some art of Stompy! <gallery> Stompy 1.png Stompy 2.png Stompy 3.png Stompy 4.png </gallery> [[Category:Robots]] [[Category:Open Source]] [[Category:K-Scale]] 7cff241cdb3456cd93e8af7602554cb0e22b7f18 0 158 983 669 2024-05-11T00:27:38Z Hatsmagee 26 buit -> bit wikitext text/x-wiki '''MIT Cheetah''' is an open-source quadropedal robot designed using low-inertia actuators. This robot is developed in the lab of Professor Sangbae Kim at MIT<ref>https://www.csail.mit.edu/news/one-giant-leap-mini-cheetah</ref>. The MIT Cheetah is known for its agility and ability to adapt to varying terrain conditions without requiring a terrain map in advance {{infobox robot | name = MIT Cheetah | organization = MIT | video_link = | cost = | height = | weight = 20 pounds | speed = Twice the speed of average human walking speed<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>,<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref> | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = Active }} == Design and Development == The MIT Cheetah displays remarkable adaptability in its functions. Despite being only 20 pounds, the robot can bend and swing its legs wide, allowing it to either walk right side up or upside down<ref>https://news.mit.edu/2019/mit-mini-cheetah-first-four-legged-robot-to-backflip-0304</ref>. This flexibility is a result of its design that prioritizes a wide range of motion. The robot is further equipped to deal with challenges posed by rough, uneven terrain. It can traverse such landscapes at a pace twice as fast as an average person's walking speed<ref>http://robotics.mit.edu/mini-cheetah-first-four-legged-robot-do-backflip</ref>. The MIT Cheetah’s design, particularly the implementation of a control system that enables agile running, has been largely driven by the "learn-by-experience model". This approach, in contrast to previous designs reliant primarily on human analytical insights, allows the robot to respond quickly to changes in the environment<ref>https://news.mit.edu/2022/3-questions-how-mit-mini-cheetah-learns-run-fast-0317</ref>. === Chips === * [https://www.st.com/en/microcontrollers-microprocessors/stm32-32-bit-arm-cortex-mcus.html STM32 32-bit Arm Cortex MCU] * [https://www.ti.com/product/DRV8323 DRV8323 3-phase smart gate driver] * [https://www.monolithicpower.com/en/ma702.html MA702 angular position measurement device] * [https://www.microchip.com/en-us/product/mcp2515 MCP2515 CAN Controller with SPI Interface] == References == <references /> [[Category:Actuators]] [[Category:Open Source]] [[Category:Robots]] c48be87863752a37917a4aa2359e11bc9936780e 5 230 984 2024-05-12T05:39:17Z 204.15.110.167 0 Created page with "Winston here from Iowa. I'm always watching to see what newer sites are going up and I just wanted to see if you would like an extra hand with getting some targeted traffic, C..." wikitext text/x-wiki Winston here from Iowa. I'm always watching to see what newer sites are going up and I just wanted to see if you would like an extra hand with getting some targeted traffic, Create custom AI bots to answer questions from visitors on your site or walk them through a sales process, create videos/images/adcopy, remove negative listings, the list goes on. I'll even shoulder 90% of the costs, dedicating my time and tools that I've created myself and bought over the years. I've been doing this for over 22 years, helped thousands of people and have loved every minute of it. There's virtually no cost on my end to do any of this for you except for my time starting at 99 a month. I don't mean to impose; I was just curious if I could lend a hand. Brief history, I've been working from home for a couple decades now and I love helping others. I'm married, have three girls and if I can provide for them by helping you and giving back by using the tools and knowledge I've built and learned over the years, I can't think of a better win-win. It amazes me that no one else is helping others quite like I do and I'd love to show you how I can help out. So, if you need any extra help in any way, please let me know either way as I value your time and don't want to pester you. PS – If I didn’t mention something you might need help with just ask, I only mentioned a handful of things to keep this brief :-) All the best, Winston Cell - 1-319-435-1790‬ My Site (w/Live Chat) - https://cutt.ly/ww91SRIU 82a02209235463885955022c253b0960a5ae40f9 0 155 985 772 2024-05-12T22:51:17Z Ben 2 wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. == MCP2515 == The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == Applications == The [https://python-can.readthedocs.io/en/stable/ python-can] library provides Controller Area Network support for Python, providing common abstractions to different hardware devices, and a suite of utilities for sending and receiving messages on a CAN bus. There is an example of an installation on [[Jetson_Orin]] here. === Pi4 === Raspberry Pi offers an easy to deploy 2-Channel Isolated CAN Bus Expansion HAT which allows to quickly integrate it to the peripheral devices. See the [https://www.waveshare.com/wiki/2-CH_CAN_HAT tutorial] for more information === Arduino === Arduino has a good support of the MCPs with many implementations of the [https://github.com/Seeed-Studio/Seeed_Arduino_CAN drivers] === MCP2515 Driver === By default the CAN bus node is supposed to acknowledge every message on the bus weather or not that node is interested in the message. However, the interference on the network can drop some bits during the communication. In the standard mode, the node would not only continuously try to re-send the unacknowledged messages, but also after a short period it would start sending error frames and then eventually go to bus-off mode and stop. This causes sever issues when the CAN network works with multiple motors. The controller has a [http://ww1.microchip.com/downloads/en/DeviceDoc/MCP2515-Stand-Alone-CAN-Controller-with-SPI-20001801J.pdf one-shot] setup that requires changes in the driver. === Wiring === Here is some suggested equipment for wiring a CAN bus: * Molex Mini-Fit Junior connectors ** [https://www.digikey.ca/en/products/detail/molex/0638190901/9655931 Crimper] ** [https://www.digikey.ca/en/products/detail/molex/0766500013/2115996 Connector kit] ** [https://www.aliexpress.us/item/3256805730106963.html Extraction tool] * CAN bus cable ** [https://www.digikey.ca/en/products/detail/igus/CF211-02-01-02/18724291 Cable] ** [https://www.digikey.com/en/products/detail/igus/CF891-07-02/21280679 Alternative Cable] * Heat shrink ** [https://www.amazon.com/Eventronic-Heat-Shrink-Tubing-Kit-3/dp/B0BVVMCY86 Tubing] == References == <references /> [[Category:Communication]] 4bd94ec196d10b06c13a0ca00c59a09ae35b4bcf 0 231 986 2024-05-13T01:39:23Z Vrtnis 21 Created page with "This is a guide for setting up and experimenting with stereo cameras in your projects. This guide is incomplete and a work in progress; you can help by expanding it! == Choo..." wikitext text/x-wiki This is a guide for setting up and experimenting with stereo cameras in your projects. This guide is incomplete and a work in progress; you can help by expanding it! == Choosing the Right Stereo Camera == In the realm of computer vision, selecting an appropriate stereo camera is fundamental. Considerations such as resolution, compatibility, and specific features like Image Signal Processing (ISP) support are paramount. For example, the [https://www.arducam.com/product/arducam-pivariety-18mp-ar1820hs-color-camera-module-for-rpi-cam-b0367/ Arducam Pivariety 18MP AR1820HS camera module] offers high resolution and is compatible with Raspberry Pi models, featuring auto exposure, auto white balance, and lens shading that are crucial for capturing high-quality images under varying lighting conditions. == Implementation and Testing == Setting up and testing stereo cameras can vary based on the project's needs. For example, streaming from a USB stereo camera to a VR headset like the Quest Pro involves addressing challenges such as latency and the processing of hand tracking data. Utilizing resources like the [https://github.com/OpenTeleVision/TeleVision TeleVision GitHub repository] can be invaluable for developers aiming to stream camera feeds efficiently, crucial for applications requiring real-time data such as virtual reality or remote operation environments. == Application Scenarios == Stereo cameras are versatile and can be adapted for numerous applications. For instance, one setup might utilize long cables for full-room scale monitoring, another for 360-degree local vision, and a third for specific stereo vision tasks. These configurations cater to the unique requirements of each application, whether it’s monitoring large spaces or creating immersive user experiences. == Computational Considerations == When deploying stereo cameras, considering the computational load is crucial. Processing two raw images from stereo pairs might seem redundant, especially if the images are similar. Techniques like using CLIP-like models for encoding can reduce the need for processing both images in depth, as these models can intuit depth from high-level semantic content, thus conserving computational resources. == Exploring Depth Sensing Techniques == Depth sensing in stereo cameras can be achieved through various technologies. While some utilize stereo disparity, others may incorporate structured light sensors for depth detection. Understanding the underlying technology is essential for optimizing the setup and ensuring efficient processing, as seen in RealSense cameras, which combine structured light sensing with stereo disparity to provide robust depth information without significant additional computational demands. 879eb32ce519bd05eee39767b76cdb7988518d8b 987 986 2024-05-13T01:47:35Z Vrtnis 21 wikitext text/x-wiki This is a guide for setting up and experimenting with stereo cameras in your projects. This guide is incomplete and a work in progress; you can help by expanding it! == Choosing the Right Stereo Camera == In the realm of computer vision, selecting an appropriate stereo camera is fundamental. Considerations such as resolution, compatibility, and specific features like Image Signal Processing (ISP) support are paramount. For example, the [https://www.arducam.com/product/arducam-pivariety-18mp-ar1820hs-color-camera-module-for-rpi-cam-b0367/ Arducam Pivariety 18MP AR1820HS camera module] offers high resolution and is compatible with Raspberry Pi models, featuring auto exposure, auto white balance, and lens shading that are crucial for capturing high-quality images under varying lighting conditions. == Implementation and Testing == Setting up and testing stereo cameras can vary based on the project's needs. For example, streaming from a USB stereo camera to a VR headset like the Quest Pro involves addressing challenges such as latency and the processing of hand tracking data. Utilizing resources like the [https://github.com/OpenTeleVision/TeleVision TeleVision GitHub repository] can be invaluable for developers aiming to stream camera feeds efficiently, crucial for applications requiring real-time data such as virtual reality or remote operation environments. == Application Scenarios == Stereo cameras are versatile and can be adapted for numerous applications. For instance, one setup might utilize long cables for full-room scale monitoring, another for 360-degree local vision, and a third for specific stereo vision tasks. These configurations cater to the unique requirements of each application, whether it’s monitoring large spaces or creating immersive user experiences. == Computational Considerations == When deploying stereo cameras, considering the computational load is crucial. Processing two raw images from stereo pairs might seem redundant, especially if the images are similar. Techniques like using CLIP-like models for encoding can reduce the need for processing both images in depth, as these models can intuit depth from high-level semantic content, thus conserving computational resources. == Exploring Depth Sensing Techniques == Depth sensing in stereo cameras can be achieved through various technologies. While some utilize stereo disparity, others may incorporate structured light sensors for depth detection. Understanding the underlying technology is essential for optimizing the setup and ensuring efficient processing, as seen in RealSense cameras, which combine structured light sensing with stereo disparity to provide robust depth information without significant additional computational demands. == Community Insights on Stereo Cameras == The community has shared varied experiences and recommendations on stereo cameras, emphasizing the practical use and applications of different models. Notably, Intel RealSense cameras seem popular among users for their robust software and integration with ROS. Despite some criticisms about the small baseline of the RealSense cameras, which can limit depth perception, alternatives like the Oak-D camera and Arducam's stereo cameras have been suggested for different needs. Oak-D is praised for its edge computing capabilities and high-quality image processing, while Arducam offers affordable options for Raspberry Pi and NVIDIA platforms. Additionally, advanced users have discussed using the ZED2 camera for its superior baseline and resolution, and have compared various models for specific needs, such as indoor testing and 3D benchmarking with Kinect and Orbbec Astra cameras. The community also highlights the importance of considering off-the-shelf depth cameras that integrate stereo computation internally to save on development time and effort. This section reflects ongoing discussions and is open for further contributions and updates from the community. e2486ca6819c8519ad48dee8b390874de57f1503 988 987 2024-05-13T01:48:17Z Vrtnis 21 wikitext text/x-wiki This is a guide for setting up and experimenting with stereo cameras in your projects. This guide is incomplete and a work in progress; you can help by expanding it! == Choosing the Right Stereo Camera == In the realm of computer vision, selecting an appropriate stereo camera is fundamental. Considerations such as resolution, compatibility, and specific features like Image Signal Processing (ISP) support are paramount. For example, the [https://www.arducam.com/product/arducam-pivariety-18mp-ar1820hs-color-camera-module-for-rpi-cam-b0367/ Arducam Pivariety 18MP AR1820HS camera module] offers high resolution and is compatible with Raspberry Pi models, featuring auto exposure, auto white balance, and lens shading that are crucial for capturing high-quality images under varying lighting conditions. == Implementation and Testing == Setting up and testing stereo cameras can vary based on the project's needs. For example, streaming from a USB stereo camera to a VR headset like the Quest Pro involves addressing challenges such as latency and the processing of hand tracking data. Utilizing resources like the [https://github.com/OpenTeleVision/TeleVision TeleVision GitHub repository] can be invaluable for developers aiming to stream camera feeds efficiently, crucial for applications requiring real-time data such as virtual reality or remote operation environments. == Application Scenarios == Stereo cameras are versatile and can be adapted for numerous applications. For instance, one setup might utilize long cables for full-room scale monitoring, another for 360-degree local vision, and a third for specific stereo vision tasks. These configurations cater to the unique requirements of each application, whether it’s monitoring large spaces or creating immersive user experiences. == Computational Considerations == When deploying stereo cameras, considering the computational load is crucial. Processing two raw images from stereo pairs might seem redundant, especially if the images are similar. Techniques like using CLIP-like models for encoding can reduce the need for processing both images in depth, as these models can intuit depth from high-level semantic content, thus conserving computational resources. == Exploring Depth Sensing Techniques == Depth sensing in stereo cameras can be achieved through various technologies. While some utilize stereo disparity, others may incorporate structured light sensors for depth detection. Understanding the underlying technology is essential for optimizing the setup and ensuring efficient processing, as seen in RealSense cameras, which combine structured light sensing with stereo disparity to provide robust depth information without significant additional computational demands. == Community Insights on Stereo Cameras == The community has shared varied experiences and recommendations on stereo cameras, emphasizing the practical use and applications of different models. Notably, Intel RealSense cameras seem popular among users for their robust software and integration with ROS. Despite some criticisms about the small baseline of the RealSense cameras, which can limit depth perception, alternatives like the Oak-D camera and Arducam's stereo cameras have been suggested for different needs. Oak-D is praised for its edge computing capabilities and high-quality image processing, while Arducam offers affordable options for Raspberry Pi and NVIDIA platforms. Additionally, advanced users have discussed using the ZED2 camera for its superior baseline and resolution, and have compared various models for specific needs, such as indoor testing and 3D benchmarking with Kinect and Orbbec Astra cameras. The community also highlights the importance of considering off-the-shelf depth cameras that integrate stereo computation internally to save on development time and effort. This section reflects ongoing discussions and is open for further contributions and updates from the community. 13033588ddc4fac7dc5e1f78a4bca591f60fdc51 989 988 2024-05-13T01:51:34Z Vrtnis 21 wikitext text/x-wiki This is a guide for setting up and experimenting with stereo cameras in your projects. This guide is incomplete and a work in progress; you can help by expanding it! == Choosing the Right Stereo Camera == In the realm of computer vision, selecting an appropriate stereo camera is fundamental. Considerations such as resolution, compatibility, and specific features like Image Signal Processing (ISP) support are paramount. For example, the [https://www.arducam.com/product/arducam-pivariety-18mp-ar1820hs-color-camera-module-for-rpi-cam-b0367/ Arducam Pivariety 18MP AR1820HS camera module] offers high resolution and is compatible with Raspberry Pi models, featuring auto exposure, auto white balance, and lens shading that are crucial for capturing high-quality images under varying lighting conditions. == Implementation and Testing == Setting up and testing stereo cameras can vary based on the project's needs. For example, streaming from a USB stereo camera to a VR headset like the Quest Pro involves addressing challenges such as latency and the processing of hand tracking data. Utilizing resources like the [https://github.com/OpenTeleVision/TeleVision TeleVision GitHub repository] can be invaluable for developers aiming to stream camera feeds efficiently, crucial for applications requiring real-time data such as virtual reality or remote operation environments. == Application Scenarios == Stereo cameras are versatile and can be adapted for numerous applications. For instance, one setup might utilize long cables for full-room scale monitoring, another for 360-degree local vision, and a third for specific stereo vision tasks. These configurations cater to the unique requirements of each application, whether it’s monitoring large spaces or creating immersive user experiences. == Computational Considerations == When deploying stereo cameras, considering the computational load is crucial. Processing two raw images from stereo pairs might seem redundant, especially if the images are similar. Techniques like using CLIP-like models for encoding can reduce the need for processing both images in depth, as these models can intuit depth from high-level semantic content, thus conserving computational resources. == Exploring Depth Sensing Techniques == Depth sensing in stereo cameras can be achieved through various technologies. While some utilize stereo disparity, others may incorporate structured light sensors for depth detection. Understanding the underlying technology is essential for optimizing the setup and ensuring efficient processing, as seen in RealSense cameras, which combine structured light sensing with stereo disparity to provide robust depth information without significant additional computational demands. == Community Insights on Stereo Cameras == The community has shared varied experiences and recommendations on stereo cameras, emphasizing the practical use and applications of different models. Notably, Intel RealSense cameras seem popular among users for their robust software and integration with ROS. Despite some criticisms about the small baseline of the RealSense cameras, which can limit depth perception, alternatives like the Oak-D camera and Arducam's stereo cameras have been suggested for different needs. Oak-D is praised for its edge computing capabilities and high-quality image processing, while Arducam offers affordable options for Raspberry Pi and NVIDIA platforms. Additionally, advanced users have discussed using the ZED2 camera for its superior baseline and resolution, and have compared various models for specific needs, such as indoor testing and 3D benchmarking with Kinect and Orbbec Astra cameras. The community also highlights the importance of considering off-the-shelf depth cameras that integrate stereo computation internally to save on development time and effort. Users have shared their experiences with various depth-sensing technologies across different robotic platforms. For example, the Unitree robotics platforms and the mini-cheetah from MIT have incorporated RealSense cameras for mapping and environmental interaction. The broader field of view afforded by using multiple cameras, as seen with Spot which employs five cameras, is advantageous for comprehensive environmental awareness. However, for specific research applications where movement is predominantly forward, a single camera might suffice. The discussion also highlights the shift towards more sophisticated sensing technologies like solid-state lidars, though cost and weight remain significant considerations. For instance, the CEREBERUS team from the SubTChallenge noted that while lidars provide more efficient depth data than multiple depth cameras, their higher cost and weight could be limiting factors depending on the robotic platform. This section reflects ongoing discussions and is open for further contributions and updates from the community. References: [1] https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9196777 [2] https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8642374 [3] https://arxiv.org/pdf/2201.07067.pdf f0f541ca4793768f23710895a39d366454e7c7f5 990 989 2024-05-13T01:52:34Z Vrtnis 21 wikitext text/x-wiki This is a guide for setting up and experimenting with stereo cameras in your projects. This guide is incomplete and a work in progress; you can help by expanding it! == Choosing the Right Stereo Camera == In the realm of computer vision, selecting an appropriate stereo camera is fundamental. Considerations such as resolution, compatibility, and specific features like Image Signal Processing (ISP) support are paramount. For example, the [https://www.arducam.com/product/arducam-pivariety-18mp-ar1820hs-color-camera-module-for-rpi-cam-b0367/ Arducam Pivariety 18MP AR1820HS camera module] offers high resolution and is compatible with Raspberry Pi models, featuring auto exposure, auto white balance, and lens shading that are crucial for capturing high-quality images under varying lighting conditions. == Implementation and Testing == Setting up and testing stereo cameras can vary based on the project's needs. For example, streaming from a USB stereo camera to a VR headset like the Quest Pro involves addressing challenges such as latency and the processing of hand tracking data. Utilizing resources like the [https://github.com/OpenTeleVision/TeleVision TeleVision GitHub repository] can be invaluable for developers aiming to stream camera feeds efficiently, crucial for applications requiring real-time data such as virtual reality or remote operation environments. == Application Scenarios == Stereo cameras are versatile and can be adapted for numerous applications. For instance, one setup might utilize long cables for full-room scale monitoring, another for 360-degree local vision, and a third for specific stereo vision tasks. These configurations cater to the unique requirements of each application, whether it’s monitoring large spaces or creating immersive user experiences. == Computational Considerations == When deploying stereo cameras, considering the computational load is crucial. Processing two raw images from stereo pairs might seem redundant, especially if the images are similar. Techniques like using CLIP-like models for encoding can reduce the need for processing both images in depth, as these models can intuit depth from high-level semantic content, thus conserving computational resources. == Exploring Depth Sensing Techniques == Depth sensing in stereo cameras can be achieved through various technologies. While some utilize stereo disparity, others may incorporate structured light sensors for depth detection. Understanding the underlying technology is essential for optimizing the setup and ensuring efficient processing, as seen in RealSense cameras, which combine structured light sensing with stereo disparity to provide robust depth information without significant additional computational demands. == Community Insights on Stereo Cameras == The community has shared varied experiences and recommendations on stereo cameras, emphasizing the practical use and applications of different models. Notably, Intel RealSense cameras seem popular among users for their robust software and integration with ROS. Despite some criticisms about the small baseline of the RealSense cameras, which can limit depth perception, alternatives like the Oak-D camera and Arducam's stereo cameras have been suggested for different needs. Oak-D is praised for its edge computing capabilities and high-quality image processing, while Arducam offers affordable options for Raspberry Pi and NVIDIA platforms. Additionally, advanced users have discussed using the ZED2 camera for its superior baseline and resolution, and have compared various models for specific needs, such as indoor testing and 3D benchmarking with Kinect and Orbbec Astra cameras. The community also highlights the importance of considering off-the-shelf depth cameras that integrate stereo computation internally to save on development time and effort. Users have shared their experiences with various depth-sensing technologies across different robotic platforms. For example, the Unitree robotics platforms and the mini-cheetah from MIT have incorporated RealSense cameras for mapping and environmental interaction. The broader field of view afforded by using multiple cameras, as seen with Spot which employs five cameras, is advantageous for comprehensive environmental awareness. However, for specific research applications where movement is predominantly forward, a single camera might suffice. The discussion also highlights the shift towards more sophisticated sensing technologies like solid-state lidars, though cost and weight remain significant considerations. For instance, the CEREBERUS team from the SubTChallenge noted that while lidars provide more efficient depth data than multiple depth cameras, their higher cost and weight could be limiting factors depending on the robotic platform. This section reflects ongoing discussions and is open for further contributions and updates from the community. References: [1] https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9196777 [2] https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8642374 [3] https://arxiv.org/pdf/2201.07067.pdf bd1f55b2f6d2c9cfac54aef19eb210921c5a6000 991 990 2024-05-13T01:53:44Z Vrtnis 21 wikitext text/x-wiki This is a guide for setting up and experimenting with stereo cameras in your projects. This guide is incomplete and a work in progress; you can help by expanding it! == Choosing the Right Stereo Camera == In the realm of computer vision, selecting an appropriate stereo camera is fundamental. Considerations such as resolution, compatibility, and specific features like Image Signal Processing (ISP) support are paramount. For example, the [https://www.arducam.com/product/arducam-pivariety-18mp-ar1820hs-color-camera-module-for-rpi-cam-b0367/ Arducam Pivariety 18MP AR1820HS camera module] offers high resolution and is compatible with Raspberry Pi models, featuring auto exposure, auto white balance, and lens shading that are crucial for capturing high-quality images under varying lighting conditions. == Implementation and Testing == Setting up and testing stereo cameras can vary based on the project's needs. For example, streaming from a USB stereo camera to a VR headset like the Quest Pro involves addressing challenges such as latency and the processing of hand tracking data. Utilizing resources like the [https://github.com/OpenTeleVision/TeleVision TeleVision GitHub repository] can be invaluable for developers aiming to stream camera feeds efficiently, crucial for applications requiring real-time data such as virtual reality or remote operation environments. == Application Scenarios == Stereo cameras are versatile and can be adapted for numerous applications. For instance, one setup might utilize long cables for full-room scale monitoring, another for 360-degree local vision, and a third for specific stereo vision tasks. These configurations cater to the unique requirements of each application, whether it’s monitoring large spaces or creating immersive user experiences. == Computational Considerations == When deploying stereo cameras, considering the computational load is crucial. Processing two raw images from stereo pairs might seem redundant, especially if the images are similar. Techniques like using CLIP-like models for encoding can reduce the need for processing both images in depth, as these models can intuit depth from high-level semantic content, thus conserving computational resources. == Exploring Depth Sensing Techniques == Depth sensing in stereo cameras can be achieved through various technologies. While some utilize stereo disparity, others may incorporate structured light sensors for depth detection. Understanding the underlying technology is essential for optimizing the setup and ensuring efficient processing, as seen in RealSense cameras, which combine structured light sensing with stereo disparity to provide robust depth information without significant additional computational demands. == Community Insights on Stereo Cameras == The community has shared varied experiences and recommendations on stereo cameras, emphasizing the practical use and applications of different models. Notably, Intel RealSense cameras seem popular among users for their robust software and integration with ROS. Despite some criticisms about the small baseline of the RealSense cameras, which can limit depth perception, alternatives like the Oak-D camera and Arducam's stereo cameras have been suggested for different needs. Oak-D is praised for its edge computing capabilities and high-quality image processing, while Arducam offers affordable options for Raspberry Pi and NVIDIA platforms. Additionally, advanced users have discussed using the ZED2 camera for its superior baseline and resolution, and have compared various models for specific needs, such as indoor testing and 3D benchmarking with Kinect and Orbbec Astra cameras. The community also highlights the importance of considering off-the-shelf depth cameras that integrate stereo computation internally to save on development time and effort. Users have shared their experiences with various depth-sensing technologies across different robotic platforms. For example, the Unitree robotics platforms and the mini-cheetah from MIT have incorporated RealSense cameras for mapping and environmental interaction. The broader field of view afforded by using multiple cameras, as seen with Spot which employs five cameras, is advantageous for comprehensive environmental awareness. However, for specific research applications where movement is predominantly forward, a single camera might suffice. Builders in the robotics community also highlight the shift towards more sophisticated sensing technologies like solid-state lidars, though cost and weight remain significant considerations. For instance, the CEREBERUS team from the SubTChallenge noted that while lidars provide more efficient depth data than multiple depth cameras, their higher cost and weight could be limiting factors depending on the robotic platform. This section reflects ongoing discussions and is open for further contributions and updates from the community. References: [1] https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9196777 [2] https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8642374 [3] https://arxiv.org/pdf/2201.07067.pdf 197e934b79d6dbc0df8a55cf483d470d1b63180e 2 232 992 2024-05-13T02:49:16Z Dennisc 27 Created page with "Dennis is a technical intern and a 2nd-year CS/Math student at CMU. {{infobox person | name = Dennis Chen | organization = [[K-Scale Labs]] | title = Technical Intern | websi..." wikitext text/x-wiki Dennis is a technical intern and a 2nd-year CS/Math student at CMU. {{infobox person | name = Dennis Chen | organization = [[K-Scale Labs]] | title = Technical Intern | website_link = https://andrew.cmu.edu/~dennisc2 }} [[Category: K-Scale Employees]] 674a6d4feb9f7b1a0d2fb07414752de4ceb48c74 0 233 997 2024-05-13T07:32:04Z Ben 2 Created page with "The G1 humanoid robot is an upcoming humanoid robot from [[Unitree]]. {{infobox robot | name = G1 | organization = [[Unitree]] | video_link = https://mp.weixin.qq.com/s/RGNVR..." wikitext text/x-wiki The G1 humanoid robot is an upcoming humanoid robot from [[Unitree]]. {{infobox robot | name = G1 | organization = [[Unitree]] | video_link = https://mp.weixin.qq.com/s/RGNVRazZqDn3y_Ijemc5Kw | cost = RMB 99,000 | height = 127 cm | weight = 35 kg | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} 8e508d43852ccba65735ef68c91327fe81ec8a21 0 139 999 722 2024-05-13T17:14:32Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Onboarding === * Watch out for an email from Gusto (our HR software), with an official offer letter and instructions on how to onboard you into our system. * Once you accept, Ben will add you to the system, after which you will have to enter your bank account information in order to be paid. === Pre-Internship Checklist === * Create a wiki account and mark yourself as an employee (you can use [[User:Ben]] as a template). You'll use your account as the main way to keep track of what you've done over the course of the internship. * Contribute an article about something you find interesting. See the [[Contributing]] guide. === What To Bring === * Bring living things (cloting, toothbrush, eyc.) * For your first day, you will need documents for your I9 authorizing you to work with us. The easiest is to just bring your passport or passport card. Alternatively, you'll need your driver's license or a federal photo ID, AND your social security card or birth certificate. === Additional Notes === * For travel expenses, please purchase your own flight and keep your receipts so that we can reimburse you later. [[Category:K-Scale]] 28a762addcd0196bffcb795f8f5d1fdcdbc798b3 0 234 1000 2024-05-13T17:49:18Z Vrtnis 21 Created page with "This guide provides insights and recommendations for choosing cost-effective 3D LiDAR systems for various applications. ==== Unitree L1 ==== This model is one of the most aff..." wikitext text/x-wiki This guide provides insights and recommendations for choosing cost-effective 3D LiDAR systems for various applications. ==== Unitree L1 ==== This model is one of the most affordable 3D LiDAR systems that might fit the budget constraints of students and hobbyists. While not perfect, it offers basic functionality needed for entry-level 3D scanning tasks. ==== Stereoscopic Cameras ==== An alternative to traditional LiDAR systems, stereoscopic cameras use dual lenses to capture spatial data, providing a 3D perception at a potentially lower cost. This technology can be a viable option for those unable to afford dedicated 3D LiDAR systems. === Understanding LiDAR Technologies === ==== Regular 2D LiDAR ==== These devices create a 2D plane of points by rotating around a vertical axis. They are generally more affordable and can offer a high sample rate, making them suitable for applications requiring planar data. ==== Stacked 2D LiDARs (Multi-layer 3D LiDAR) ==== Often marketed as 3D LiDAR, these are essentially multiple 2D LiDARs arranged vertically at different angles, rotating around the same axis. They tend to be expensive and do not produce a dense 3D point cloud unless moved in space, making them less ideal for static applications. ==== 3D Scanners ==== These systems use a 2D LiDAR tilted 90 degrees, combined with a secondary mechanism that rotates the device around its longitudinal axis. While they generate detailed 3D point clouds, they do so slowly and are not suited for real-time applications. === Budget-Friendly Alternatives and DIY Solutions === ==== Smartphone Depth Sensors ==== For those with access to newer smartphones, utilizing built-in depth sensors can be a cost-effective way to gather 3D data. These devices, like the iPhone's TrueDepth camera, can generate usable 2D depth maps or 3D point clouds. ==== DIY 3D Scanners ==== Constructing a 3D scanner from components such as a 2D LiDAR, a Raspberry Pi, and a camera can be an educational and affordable project. This approach requires more technical expertise but allows for customization and potentially lower costs. Choosing the right LiDAR system depends on the specific needs of the project and the available budget. For projects where real-time data is not critical, alternatives like stereoscopic cameras or DIY solutions might be adequate. However, for more demanding applications, investing in a higher-quality 3D LiDAR may be necessary. e016c33fb7af30f9439dc172af8948408ff03dbe 1002 1000 2024-05-13T17:52:13Z Vrtnis 21 wikitext text/x-wiki This guide provides insights and recommendations for choosing cost-effective 3D LiDAR systems for various applications. This guide is incomplete and a work in progress; you can help by expanding it! ==== Unitree L1 ==== This model is one of the most affordable 3D LiDAR systems that might fit the budget constraints of students and hobbyists. While not perfect, it offers basic functionality needed for entry-level 3D scanning tasks. ==== Stereoscopic Cameras ==== An alternative to traditional LiDAR systems, stereoscopic cameras use dual lenses to capture spatial data, providing a 3D perception at a potentially lower cost. This technology can be a viable option for those unable to afford dedicated 3D LiDAR systems. === Understanding LiDAR Technologies === ==== Regular 2D LiDAR ==== These devices create a 2D plane of points by rotating around a vertical axis. They are generally more affordable and can offer a high sample rate, making them suitable for applications requiring planar data. ==== Stacked 2D LiDARs (Multi-layer 3D LiDAR) ==== Often marketed as 3D LiDAR, these are essentially multiple 2D LiDARs arranged vertically at different angles, rotating around the same axis. They tend to be expensive and do not produce a dense 3D point cloud unless moved in space, making them less ideal for static applications. ==== 3D Scanners ==== These systems use a 2D LiDAR tilted 90 degrees, combined with a secondary mechanism that rotates the device around its longitudinal axis. While they generate detailed 3D point clouds, they do so slowly and are not suited for real-time applications. === Budget-Friendly Alternatives and DIY Solutions === ==== Smartphone Depth Sensors ==== For those with access to newer smartphones, utilizing built-in depth sensors can be a cost-effective way to gather 3D data. These devices, like the iPhone's TrueDepth camera, can generate usable 2D depth maps or 3D point clouds. ==== DIY 3D Scanners ==== Constructing a 3D scanner from components such as a 2D LiDAR, a Raspberry Pi, and a camera can be an educational and affordable project. This approach requires more technical expertise but allows for customization and potentially lower costs. Choosing the right LiDAR system depends on the specific needs of the project and the available budget. For projects where real-time data is not critical, alternatives like stereoscopic cameras or DIY solutions might be adequate. However, for more demanding applications, investing in a higher-quality 3D LiDAR may be necessary. dfbb71c6b56c886ad9850df12c55d1e20c0d1170 0 235 1001 2024-05-13T17:49:31Z Matt 16 add candaddy page wikitext text/x-wiki CANdaddy info here 89d9bfea9212efe08992b78609fc3e6b68f18936 0 236 1003 2024-05-13T18:13:22Z Vrtnis 21 Created page with " Haptic technology in robotics has transformed the landscape of remote operations, offering a more nuanced and precise control that mimics human touch. This technological adv..." wikitext text/x-wiki Haptic technology in robotics has transformed the landscape of remote operations, offering a more nuanced and precise control that mimics human touch. This technological advancement allows operators to control robots from afar, feeling what the robots feel as if they were physically present, which dramatically improves the operator's ability to perform complex tasks. eb71bef2b8ae56965544398944d28afe2b853568 1004 1003 2024-05-13T18:13:57Z Vrtnis 21 wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! Haptic technology in robotics has transformed the landscape of remote operations, offering a more nuanced and precise control that mimics human touch. This technological advancement allows operators to control robots from afar, feeling what the robots feel as if they were physically present, which dramatically improves the operator's ability to perform complex tasks. e44ec013de73a3c5766137efbb1a7f216ef1999c 1005 1004 2024-05-13T18:52:04Z Vrtnis 21 wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! Haptic technology in robotics has transformed the landscape of remote operations, offering a more nuanced and precise control that mimics human touch. This technological advancement allows operators to control robots from afar, feeling what the robots feel as if they were physically present, which dramatically improves the operator's ability to perform complex tasks. * '''HaptX Gloves and the Tactile Telerobot''': Developed in collaboration with the Converge Robotics Group, the HaptX Gloves are part of the innovative Tactile Telerobot system. This setup enables precise, intuitive control over robotic hands from thousands of miles away, with the gloves providing real-time tactile sensations that simulate actual contact with various objects. . ===Improving Remote Interaction === The haptic gloves are designed to reduce latency and increase the accuracy of feedback, making remote operations feel more immediate and intuitive. For example, a notable test involved using these gloves to type a message on a keyboard situated over 5,000 miles away, demonstrating the critical role of tactile feedback in performing everyday tasks remotely. d12ea3ab97d7a1ddd2a6a25625bba2c52bc92f85 1006 1005 2024-05-13T18:54:09Z Vrtnis 21 wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! Haptic technology in robotics has transformed the landscape of remote operations, offering a more nuanced and precise control that mimics human touch. This technological advancement allows operators to control robots from afar, feeling what the robots feel as if they were physically present, which dramatically improves the operator's ability to perform complex tasks. === HaptX Gloves and the Tactile Telerobot === Developed in collaboration with the Converge Robotics Group, the HaptX Gloves are part of the innovative Tactile Telerobot system. This setup enables precise, intuitive control over robotic hands from thousands of miles away, with the gloves providing real-time tactile sensations that simulate actual contact with various objects. ===Improving Remote Interaction === The haptic gloves are designed to reduce latency and increase the accuracy of feedback, making remote operations feel more immediate and intuitive. For example, a notable test involved using these gloves to type a message on a keyboard situated over 5,000 miles away, demonstrating the critical role of tactile feedback in performing everyday tasks remotely. 017fd6cba94260e8b065f75a42a9b311c3664f08 1007 1006 2024-05-13T18:57:52Z Vrtnis 21 wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! __TOC__ Haptic technology in robotics has transformed the landscape of remote operations, offering a more nuanced and precise control that mimics human touch. This technological advancement allows operators to control robots from afar, feeling what the robots feel as if they were physically present, which dramatically improves the operator's ability to perform complex tasks. === HaptX Gloves and the Tactile Telerobot === Developed in collaboration with the Converge Robotics Group, the HaptX Gloves are part of the innovative Tactile Telerobot system. This setup enables precise, intuitive control over robotic hands from thousands of miles away, with the gloves providing real-time tactile sensations that simulate actual contact with various objects. Among the leading developments in this area is the Shadow Dexterous Hand, developed by The Shadow Robot Company in London. This humanoid robot hand is comparable to the average human hand in size and shape and exceeds the human hand in terms of mechanical flexibility and control options. ===Design and Functionality=== The Shadow Dexterous Hand boasts 24 joints and 20 degrees of freedom, surpassing the human hand's complexity. It is designed to mimic the range of movement of a typical human, featuring a sophisticated joint structure in its fingers and thumb: * Each of the four fingers is equipped with two one-axis joints linking the phalanges and one universal joint at the metacarpal. * The little finger includes an additional one-axis joint for enhanced palm curl movements. * The thumb features a one-axis joint for distal movements, a universal joint at the metacarpal, and another one-axis joint to aid palm curling. * The wrist consists of two joints that provide flexion/extension and adduction/abduction movements. The Shadow Hand is available in both electric motor-driven and pneumatic muscle-driven models. The electric version operates with DC motors located in the forearm, while the pneumatic version uses antagonistic pairs of air muscles for movement. Each model is equipped with Hall effect sensors in every joint for precise positional feedback. Additional tactile sensing capabilities range from basic pressure sensors to advanced multimodal tactile sensors like the BioTac from Syntouch Inc., enhancing the hand's sensory feedback to mimic human touch and interaction closely. ===Software and Simulation=== Control and simulation of the Shadow Hand are facilitated through the Robot Operating System (ROS), which provides tools for configuration, calibration, and simulation. This integration allows for extensive testing and development in a virtual environment, aiding in the hand's continuous improvement and adaptation for various applications. d45efe9bf871cbb56a76f119e567bd5059401f2f 0 237 1008 2024-05-13T22:16:53Z Budzianowski 19 Created page with "We want to build a simplified conversational pipeline that will allow to talk and give orders to Stompy '''on device'''. We also believe that this could be long-term wise a st..." wikitext text/x-wiki We want to build a simplified conversational pipeline that will allow to talk and give orders to Stompy '''on device'''. We also believe that this could be long-term wise a standalone product High level observations: # ASR is 'solved' for non-noisy single user speech, multi-speaker noisy environment with intelligent VAD is a hard problem # LLMs are 'solved' with 3B models being smart enough to handle dealing with orders and having laid back convo # TTS is 'solved' for non-conversational (reading news, book) use-cases with large models. Having 'intelligent' fast voice is an open problem # Building a demo is simple, creating a setup that is robust to the noise and edge-cases is not ==Pipeline parts== ===ASR=== # Latency - most of the system in the proper setup can get below <100ms but knowing when to stop to listen is a VAD problem # VAD - Voice Activity Detection (fundamental to conversational application), benchmarked typically with https://github.com/snakers4/silero-vad, most of the systems are very simple and super stupid (ML a'la 2015) # Multispeaker - most of models are clueless when it comes to many speakers environment # Barge-in - an option to interrupt the system, in any noisy environment completely disrupts the system, none of the existing ASR systems are trained to deal with this properly (ASR doesn't understand the context) # Noise - in-house data are fundamental to fine-tune big model to the specific environment (20 hours are often enough to ground on the use-case) # Keyword detection - models that focus on only spotting one word (Alexa, Siri) # Best conversational production systems - only few: Google, RIVA, Deepgram ===LLM=== # Vision support starts to be available off-the-shelf ===TTS=== # Quality - conversational vs reading abilities, most models don't really understand the context # Latency - most of open-source model are either fast or slow # Streaming option - only now there are first model truly streaming-oriented ==Current Setup== Building on the Jetson through tight Docker compose module due to multiple conflicting requirements. Dusty created a fantastic repository of many e2e setups at https://www.jetson-ai-lab.com/tutorial-intro.html However, typically ASR and TTS are based on RIVA which is not the setup you would want to keep/support. The building process for each model requires a lot of small changes to the setup. Currently, having an e2e setup requires composing different images into a pipeline loop. 0. Audio<br> The Linux interaction between ALSA, Pulsaudio, sudo vs user is ironically the most demanding part. See https://0pointer.de/blog/projects/guide-to-sound-apis.html or https://www.reddit.com/r/archlinux/comments/ae67oa/lets_talk_about_how_the_linuxarch_sound/. The most stable approach is to rely on PyAudio to handle dmix issues and often changes to the hardware availability. 1. ASR<br> * Whisper is good but nothing close to production-ready ASR system * RIVA is the best option (production ready, reliably) however requires licensing per GPU which is not a good model long term-wse * Any other options are either bad or not production ready 2. LLM<br> * Rely on ollama and NanoLLM which supports latests SLMs (below 7B) * Fast and well behaved under onnxruntime library 3. TTS<br> * FastPitch (RIVA) - fast but old-style and poor quality * StylTTS2 - really good quality, requires a big clean up to make it TensorRT ready, currently 200ms on A100 * xTTS ok quality but slow even with TensorRT (dropped) * PiperTTS (old VITS model) - fast but old-style and poor quality ==Long-term bets== # E2E 625a4a5f81d764dce16b32aa0f0701d753426c67 1009 1008 2024-05-13T22:17:09Z Budzianowski 19 wikitext text/x-wiki We want to build a simplified conversational pipeline that will allow to talk and give orders to Stompy '''on device'''. We also believe that this could be long-term wise a standalone product High level observations: # ASR is 'solved' for non-noisy single user speech, multi-speaker noisy environment with intelligent VAD is a hard problem # LLMs are 'solved' with 3B models being smart enough to handle dealing with orders and having laid back convo # TTS is 'solved' for non-conversational (reading news, book) use-cases with large models. Having 'intelligent' fast voice is an open problem # Building a demo is simple, creating a setup that is robust to the noise and edge-cases is not ==Pipeline parts== ===ASR=== # Latency - most of the system in the proper setup can get below <100ms but knowing when to stop to listen is a VAD problem # VAD - Voice Activity Detection (fundamental to conversational application), benchmarked typically with https://github.com/snakers4/silero-vad, most of the systems are very simple and super stupid (ML a'la 2015) # Multispeaker - most of models are clueless when it comes to many speakers environment # Barge-in - an option to interrupt the system, in any noisy environment completely disrupts the system, none of the existing ASR systems are trained to deal with this properly (ASR doesn't understand the context) # Noise - in-house data are fundamental to fine-tune big model to the specific environment (20 hours are often enough to ground on the use-case) # Keyword detection - models that focus on only spotting one word (Alexa, Siri) # Best conversational production systems - only few: Google, RIVA, Deepgram ===LLM=== # Vision support starts to be available off-the-shelf ===TTS=== # Quality - conversational vs reading abilities, most models don't really understand the context # Latency - most of open-source model are either fast or slow # Streaming option - only now there are first model truly streaming-oriented ==Current Setup== Building on the Jetson through tight Docker compose module due to multiple conflicting requirements. Dusty created a fantastic repository of many e2e setups at https://www.jetson-ai-lab.com/tutorial-intro.html However, typically ASR and TTS are based on RIVA which is not the setup you would want to keep/support. The building process for each model requires a lot of small changes to the setup. Currently, having an e2e setup requires composing different images into a pipeline loop. 0. Audio<br> The Linux interaction between ALSA, Pulsaudio, sudo vs user is ironically the most demanding part. See https://0pointer.de/blog/projects/guide-to-sound-apis.html or https://www.reddit.com/r/archlinux/comments/ae67oa/lets_talk_about_how_the_linuxarch_sound/. The most stable approach is to rely on PyAudio to handle dmix issues and often changes to the hardware availability. 1. ASR<br> * Whisper is good but nothing close to production-ready ASR system * RIVA is the best option (production ready, reliably) however requires licensing per GPU which is not a good model long term-wse * Any other options are either bad or not production ready 2. LLM<br> * Rely on ollama and NanoLLM which supports latests SLMs (below 7B) * Fast and well behaved under onnxruntime library 3. TTS<br> * FastPitch (RIVA) - fast but old-style and poor quality * StylTTS2 - really good quality, requires a big clean up to make it TensorRT ready, currently 200ms on A100 * xTTS ok quality but slow even with TensorRT (dropped) * PiperTTS (old VITS model) - fast but old-style and poor quality ==Long-term bets== # E2E [[Category: Firmware]] 53297fa72d22c084a22b9bd65557b23b431c7b0b 1010 1009 2024-05-13T22:44:48Z Budzianowski 19 wikitext text/x-wiki We want to build a simplified conversational pipeline that will allow to talk and give orders to Stompy '''on device'''. We also believe that this could be long-term wise a standalone product High level observations: # ASR is 'solved' for non-noisy single user speech, multi-speaker noisy environment with intelligent VAD is a hard problem # LLMs are 'solved' with 3B models being smart enough to handle dealing with orders and having laid back convo # TTS is 'solved' for non-conversational (reading news, book) use-cases with large models. Having 'intelligent' fast voice is an open problem # Building a demo is simple, creating a setup that is robust to the noise and edge-cases is not ==Pipeline parts== ===ASR=== # Latency - most of the system in the proper setup can get below <100ms but knowing when to stop to listen is a VAD problem # VAD - Voice Activity Detection (fundamental to conversational application), benchmarked typically with https://github.com/snakers4/silero-vad, most of the systems are very simple and super stupid (ML a'la 2015) # Multispeaker - most of models are clueless when it comes to many speakers environment # Barge-in - an option to interrupt the system, in any noisy environment completely disrupts the system, none of the existing ASR systems are trained to deal with this properly (ASR doesn't understand the context) # Noise - in-house data are fundamental to fine-tune big model to the specific environment (20 hours are often enough to ground on the use-case) # Keyword detection - models that focus on only spotting one word (Alexa, Siri) # Best conversational production systems - only few: Google, RIVA, Deepgram ===LLM=== # Vision support starts to be available off-the-shelf ===TTS=== # Quality - conversational vs reading abilities, most models don't really understand the context # Latency - most of open-source model are either fast or slow # Streaming option - only now there are first model truly streaming-oriented ==Current Setup== Building on the Jetson through tight Docker compose module due to multiple conflicting requirements. Dusty created a fantastic repository of many e2e setups at https://www.jetson-ai-lab.com/tutorial-intro.html However, typically ASR and TTS are based on RIVA which is not the setup you would want to keep/support. The building process for each model requires a lot of small changes to the setup. Currently, having an e2e setup requires composing different images into a pipeline loop. 0. Audio<br> The Linux interaction between ALSA, Pulsaudio, sudo vs user is ironically the most demanding part. See https://0pointer.de/blog/projects/guide-to-sound-apis.html or https://www.reddit.com/r/archlinux/comments/ae67oa/lets_talk_about_how_the_linuxarch_sound/. The most stable approach is to rely on PyAudio to handle dmix issues and often changes to the hardware availability. 1. ASR<br> * Whisper is good but nothing close to production-ready ASR system * RIVA is the best option (production ready, reliably) however requires licensing per GPU which is not a good model long term-wse * Any other options are either bad or not production ready 2. LLM<br> * Rely on ollama and NanoLLM which supports latests SLMs (below 7B) * Fast and well behaved under onnxruntime library 3. TTS<br> * FastPitch (RIVA) - fast but old-style and poor quality * [https://github.com/yl4579/StyleTTS2 StyleTTS2] - really good quality, requires a big clean up to make it TensorRT ready, currently 200ms on A100 * xTTS ok quality but slow even with TensorRT (dropped) * PiperTTS (old VITS model) - fast but old-style and poor quality ==Long-term bets== # E2E [[Category: Firmware]] 26d24f6715766da2e10266d23b8e6dde8eed9d5b 0 234 1011 1002 2024-05-14T17:22:25Z Vrtnis 21 wikitext text/x-wiki This guide provides insights and recommendations for choosing cost-effective 3D LiDAR systems for various applications. This guide is incomplete and a work in progress; you can help by expanding it! ==== Unitree L1 ==== This model is one of the most affordable 3D LiDAR systems that might fit the budget constraints of students and hobbyists. While not perfect, it offers basic functionality needed for entry-level 3D scanning tasks. ==== Stereoscopic Cameras ==== An alternative to traditional LiDAR systems, stereoscopic cameras use dual lenses to capture spatial data, providing a 3D perception at a potentially lower cost. This technology can be a viable option for those unable to afford dedicated 3D LiDAR systems. === Understanding LiDAR Technologies === ==== Regular 2D LiDAR ==== These devices create a 2D plane of points by rotating around a vertical axis. They are generally more affordable and can offer a high sample rate, making them suitable for applications requiring planar data. ==== Stacked 2D LiDARs (Multi-layer 3D LiDAR) ==== Often marketed as 3D LiDAR, these are essentially multiple 2D LiDARs arranged vertically at different angles, rotating around the same axis. They tend to be expensive and do not produce a dense 3D point cloud unless moved in space, making them less ideal for static applications. ==== 3D Scanners ==== These systems use a 2D LiDAR tilted 90 degrees, combined with a secondary mechanism that rotates the device around its longitudinal axis. While they generate detailed 3D point clouds, they do so slowly and are not suited for real-time applications. === Budget-Friendly Alternatives and DIY Solutions === ==== Smartphone Depth Sensors ==== For those with access to newer smartphones, utilizing built-in depth sensors can be a cost-effective way to gather 3D data. These devices, like the iPhone's TrueDepth camera, can generate usable 2D depth maps or 3D point clouds. ==== DIY 3D Scanners ==== Constructing a 3D scanner from components such as a 2D LiDAR, a Raspberry Pi, and a camera can be an educational and affordable project. This approach requires more technical expertise but allows for customization and potentially lower costs. Choosing the right LiDAR system depends on the specific needs of the project and the available budget. For projects where real-time data is not critical, alternatives like stereoscopic cameras or DIY solutions might be adequate. However, for more demanding applications, investing in a higher-quality 3D LiDAR may be necessary. {| class="wikitable" |+ ! Product !! Brand !! SKU !! Approx Price !! Description |- | RPlidar S2 360° Laser Scanner (30 m) || SLAMTEC || RB-Rpk-20 || $399.00 || Offers 30 meters radius ranging distance. Adopts the low power infrared laser light. Provides excellent ranging accuracy. Works well both in an indoor and outdoor environment. Proof Level: IP65 |- | Hokuyo UST-10LX Scanning Laser Rangefinder || HOKUYO || RB-Hok-24 || $1,670.00 || Small, accurate, high-speed scanning laser range finder. Obstacle detection and localization of autonomous robots. Detection distance (maximum): 30m. Light source: Laser semiconductor (905nm). Scan speed: 25ms (Motor speed 2400rpm) |- | RPLidar A1M8 - 360 Degree Laser Scanner Development Kit || SLAMTEC || RB-Rpk-03 || $85.00 || 360° Omnidirectional Laser Scan. 5.5 - 10Hz Adaptive Scan Frequency. Sample Frequency: 4000 - 8000Hz. Distance range: 0.15 - 12m. Works With SLAMAIg |- | Slamtec RPLIDAR A3 360° Laser Scanner (25 m) || SLAMTEC || RB-Rpk-07 || $599.00 || White object: 25 m, Dark object: 10 m (Enhanced mode). White object: 20 m, Dark object: TBD (Outdoor mode). Sample Rate: 16000 - 10000 times per second. Scan Rate: 10-20 Hz. Angular Resolution: 0.225°. Supports former SDK protocols |- | Slamtec Mapper Pro M2M2 360° Laser Mapping Sensor (40 m) || SLAMTEC || RB-Rpk-15 || $699.00 || Features a 40 m laser mapping sensor. Provides large scenarios and high-quality mapping. Offers integrating map building and real-time localization and navigation. Offers tilting compensation and fast-moving |- | LIDAR-Lite 3 Laser Rangefinder || GARMIN || RB-Pli-06 || $124.95 || Compact 48mm x 40mm x 20mm module with 40m measuring range. Signal Processing Improvements offer 5X Faster Measurement Speeds. Improved I2C Communications and assignable I2C Addressing. Great for drones, robotics and other demanding applications |- | Benewake TF03 LIDAR LED Rangefinder IP67 (100 m) || BENEWAKE || RB-Ben-12 || $218.46 || Features excellent performance and a compact size. Upgrades more than ten key indicators and opens multiple expansion functions. Is used in the fields of intelligent transportation and industrial safety warning. Provides an industrial grade, long-distance of 100 m |- | RPLIDAR A2M12 360° Laser Range Scanner || SLAMTEC || RB-Rpk-22 || $229.00 || Offers a low-cost 360° Laser Range Scanner. Comes with a rotation speed detection and adaptive system. Features low noise brushless motor. Adopts the low-power infrared laser light. Can run smoothly without any noise |- | RPLIDAR A2M8 360° Laser Scanner (DISCONTINUED) || SLAMTEC || RB-Rpk-02 || $319.00 || Low Cost 360 Degree Laser Range Scanner. Sample Frequency: 2000 - 8000 Hz. Scan Rate: 5 - 15Hz. Distance Range: 0.15 - 12m. Angular Resolution: 0.45 - 1.35°. Barrel connector available Small Barrel Connector Assembly 0.7x2.35 mm (Needs Soldering) |- | LIDAR-Lite 3 Laser Rangefinder High Performance (LLV3HP) || GARMIN || RB-Pli-17 || $149.05 || Compact 53mm x 33mm x 24mm module with 40m measuring range. Sturdy IPX7-rated housing for drone, robot or unmanned vehicle applications. User configurability allows adjustment between accuracy, operating range and measurement time. Communicates via I2C and PWM. Low power consumption; requires less than 85 milliamps during acquisition |- | LD19 D300 LiDAR Developer Kit, 360 DToF Laser Scanner, Supports ROS1/2, Raspberry Pi & Jetson Nano || HIWONDER || RM-HIWO-040 || $98.99 || Long Service Time of 10,000 hours. Resistance to Strong Light of 30,000 lux. Distance Measurement up to a Radius of 12 m. All-round Laser Scanning of 360 degrees |- | Benewake TFMINI Plus Micro LIDAR Module UART/I2C (12 m) || BENEWAKE || RB-Ben-09 || $49.90 || Features a cost-effective LiDAR. Offers a small-size and low power consumption. Improves the frame rate. Introduces IP65 enclosures. Requires TTL-USB converter to change between UART and I2C |- | RPLIDAR S1 360° Laser Scanner (40 m) || SLAMTEC || RB-Rpk-13 || $649.00 || Can take up to 9200 samples of laser ranging per second. Offers a high rotation speed. Equipped with SLAMTEC patented OPTMAG technology |- | Benewake TF-Luna 8m LiDAR Distance Sensor || BENEWAKE || RB-Ben-17 || $19.98 || Features a low-cost ranging LiDAR module. Has an operating range of 0.2 - 8 m. Gives unique optical and electrical design. Provides highly stable, accurate, sensitive range detection. Supports power saving mode for power-sensitive applications. Suitable for scenarios with strict load requirements |- | Slamtec RPlidar S3 360° Laser Scanner (40 m) || SLAMTEC || RB-Rpk-29 || $549.00 || Offers a next-generation low-cost 360-degree 2D laser scanner. Smaller size, better performance. Comes with a rotation speed detection and adaptive system. Performs a 2D 360-degree scan within a 40-meter range. Low reflectivity at 15 meters measurement radius. Ideal for both outdoor and indoor environments. Compatible with the SLAMTEC ecosystem |- | YDLIDAR TG30 360° Laser Scanner (30 m) || YDLIDAR || RB-Ydl-09 || $399.00 || Offers 360-degree omnidirectional scanning. Compact structure, suitable for integration. Provides TOF (Time of Flight) ranging technology. Enclosed housing and IP65 proof level. Equipped with related optics, electricity, and algorithm |- | YDLIDAR G2 360° Laser Scanner || YDLIDAR || RB-Ydl-05 || $159.00 || Features a 360° Omnidirectional scanning range. Offers small distance error, stable performance and high accuracy. Has a 5-12 Hz adaptive scanning frequency |- | Benewake TF02-Pro LIDAR LED Rangefinder IP65 (40m) || BENEWAKE || RB-Ben-14 || $86.39 || Features a single-point ranging LiDAR. Can achieve stable, accuracy, sensitive and high-frequency range detection. Offers a sensor range up to 40 meters. Has an ambient light resistance up to 100 Klux |} 00b24fa06ed93f8df6481a771ae4ddf2da094084 1012 1011 2024-05-14T17:28:06Z Vrtnis 21 wikitext text/x-wiki This guide provides insights and recommendations for choosing cost-effective 3D LiDAR systems for various applications. This guide is incomplete and a work in progress; you can help by expanding it! ==== Unitree L1 ==== This model is one of the most affordable 3D LiDAR systems that might fit the budget constraints of students and hobbyists. While not perfect, it offers basic functionality needed for entry-level 3D scanning tasks. ==== Stereoscopic Cameras ==== An alternative to traditional LiDAR systems, stereoscopic cameras use dual lenses to capture spatial data, providing a 3D perception at a potentially lower cost. This technology can be a viable option for those unable to afford dedicated 3D LiDAR systems. === Understanding LiDAR Technologies === ==== Regular 2D LiDAR ==== These devices create a 2D plane of points by rotating around a vertical axis. They are generally more affordable and can offer a high sample rate, making them suitable for applications requiring planar data. ==== Stacked 2D LiDARs (Multi-layer 3D LiDAR) ==== Often marketed as 3D LiDAR, these are essentially multiple 2D LiDARs arranged vertically at different angles, rotating around the same axis. They tend to be expensive and do not produce a dense 3D point cloud unless moved in space, making them less ideal for static applications. ==== 3D Scanners ==== These systems use a 2D LiDAR tilted 90 degrees, combined with a secondary mechanism that rotates the device around its longitudinal axis. While they generate detailed 3D point clouds, they do so slowly and are not suited for real-time applications. === Budget-Friendly Alternatives and DIY Solutions === ==== Smartphone Depth Sensors ==== For those with access to newer smartphones, utilizing built-in depth sensors can be a cost-effective way to gather 3D data. These devices, like the iPhone's TrueDepth camera, can generate usable 2D depth maps or 3D point clouds. ==== DIY 3D Scanners ==== Constructing a 3D scanner from components such as a 2D LiDAR, a Raspberry Pi, and a camera can be an educational and affordable project. This approach requires more technical expertise but allows for customization and potentially lower costs. Choosing the right LiDAR system depends on the specific needs of the project and the available budget. For projects where real-time data is not critical, alternatives like stereoscopic cameras or DIY solutions might be adequate. However, for more demanding applications, investing in a higher-quality 3D LiDAR may be necessary. {| class="wikitable" |+ ! Product !! Brand !! Approx Price !! Description |- | RPlidar S2 360° Laser Scanner (30 m) || SLAMTEC || $399.00 || Offers 30 meters radius ranging distance. Adopts the low power infrared laser light. Provides excellent ranging accuracy. Works well both in an indoor and outdoor environment. Proof Level: IP65 |- | Hokuyo UST-10LX Scanning Laser Rangefinder || HOKUYO || $1,670.00 || Small, accurate, high-speed scanning laser range finder. Obstacle detection and localization of autonomous robots. Detection distance (maximum): 30m. Light source: Laser semiconductor (905nm). Scan speed: 25ms (Motor speed 2400rpm) |- | RPLidar A1M8 - 360 Degree Laser Scanner Development Kit || SLAMTEC || $85.00 || 360° Omnidirectional Laser Scan. 5.5 - 10Hz Adaptive Scan Frequency. Sample Frequency: 4000 - 8000Hz. Distance range: 0.15 - 12m. Works With SLAMAIg |- | Slamtec RPLIDAR A3 360° Laser Scanner (25 m) || SLAMTEC || $599.00 || White object: 25 m, Dark object: 10 m (Enhanced mode). White object: 20 m, Dark object: TBD (Outdoor mode). Sample Rate: 16000 - 10000 times per second. Scan Rate: 10-20 Hz. Angular Resolution: 0.225°. Supports former SDK protocols |- | Slamtec Mapper Pro M2M2 360° Laser Mapping Sensor (40 m) || SLAMTEC || $699.00 || Features a 40 m laser mapping sensor. Provides large scenarios and high-quality mapping. Offers integrating map building and real-time localization and navigation. Offers tilting compensation and fast-moving |- | LIDAR-Lite 3 Laser Rangefinder || GARMIN || $124.95 || Compact 48mm x 40mm x 20mm module with 40m measuring range. Signal Processing Improvements offer 5X Faster Measurement Speeds. Improved I2C Communications and assignable I2C Addressing. Great for drones, robotics and other demanding applications |- | Benewake TF03 LIDAR LED Rangefinder IP67 (100 m) || BENEWAKE || $218.46 || Features excellent performance and a compact size. Upgrades more than ten key indicators and opens multiple expansion functions. Is used in the fields of intelligent transportation and industrial safety warning. Provides an industrial grade, long-distance of 100 m |- | RPLIDAR A2M12 360° Laser Range Scanner || SLAMTEC || $229.00 || Offers a low-cost 360° Laser Range Scanner. Comes with a rotation speed detection and adaptive system. Features low noise brushless motor. Adopts the low-power infrared laser light. Can run smoothly without any noise |- | RPLIDAR A2M8 360° Laser Scanner (DISCONTINUED) || SLAMTEC || $319.00 || Low Cost 360 Degree Laser Range Scanner. Sample Frequency: 2000 - 8000 Hz. Scan Rate: 5 - 15Hz. Distance Range: 0.15 - 12m. Angular Resolution: 0.45 - 1.35°. Barrel connector available Small Barrel Connector Assembly 0.7x2.35 mm (Needs Soldering) |- | LIDAR-Lite 3 Laser Rangefinder High Performance (LLV3HP) || GARMIN || $149.05 || Compact 53mm x 33mm x 24mm module with 40m measuring range. Sturdy IPX7-rated housing for drone, robot or unmanned vehicle applications. User configurability allows adjustment between accuracy, operating range and measurement time. Communicates via I2C and PWM. Low power consumption; requires less than 85 milliamps during acquisition |- | LD19 D300 LiDAR Developer Kit, 360 DToF Laser Scanner, Supports ROS1/2, Raspberry Pi & Jetson Nano || HIWONDER || $98.99 || Long Service Time of 10,000 hours. Resistance to Strong Light of 30,000 lux. Distance Measurement up to a Radius of 12 m. All-round Laser Scanning of 360 degrees |- | Benewake TFMINI Plus Micro LIDAR Module UART/I2C (12 m) || BENEWAKE || $49.90 || Features a cost-effective LiDAR. Offers a small-size and low power consumption. Improves the frame rate. Introduces IP65 enclosures. Requires TTL-USB converter to change between UART and I2C |- | RPLIDAR S1 360° Laser Scanner (40 m) || SLAMTEC || $649.00 || Can take up to 9200 samples of laser ranging per second. Offers a high rotation speed. Equipped with SLAMTEC patented OPTMAG technology |- | Benewake TF-Luna 8m LiDAR Distance Sensor || BENEWAKE || $19.98 || Features a low-cost ranging LiDAR module. Has an operating range of 0.2 - 8 m. Gives unique optical and electrical design. Provides highly stable, accurate, sensitive range detection. Supports power saving mode for power-sensitive applications. Suitable for scenarios with strict load requirements |- | Slamtec RPlidar S3 360° Laser Scanner (40 m) || SLAMTEC || $549.00 || Offers a next-generation low-cost 360-degree 2D laser scanner. Smaller size, better performance. Comes with a rotation speed detection and adaptive system. Performs a 2D 360-degree scan within a 40-meter range. Low reflectivity at 15 meters measurement radius. Ideal for both outdoor and indoor environments. Compatible with the SLAMTEC ecosystem |- | YDLIDAR TG30 360° Laser Scanner (30 m) || YDLIDAR || $399.00 || Offers 360-degree omnidirectional scanning. Compact structure, suitable for integration. Provides TOF (Time of Flight) ranging technology. Enclosed housing and IP65 proof level. Equipped with related optics, electricity, and algorithm |- | YDLIDAR G2 360° Laser Scanner || YDLIDAR || $159.00 || Features a 360° Omnidirectional scanning range. Offers small distance error, stable performance and high accuracy. Has a 5-12 Hz adaptive scanning frequency |- | Benewake TF02-Pro LIDAR LED Rangefinder IP65 (40m) || BENEWAKE || $86.39 || Features a single-point ranging LiDAR. Can achieve stable, accuracy, sensitive and high-frequency range detection. Offers a sensor range up to 40 meters. Has an ambient light resistance up to 100 Klux |} fe3724ba4ea0a57290524d92496700097cbf78f3 0 238 1013 2024-05-14T17:42:54Z Vrtnis 21 Created page with "[https://global.honda/products/robotics.html Honda Robotics] is a robotics division of Honda based in Japan. They are known for developing the advanced humanoid robot [[ASIMO]..." wikitext text/x-wiki [https://global.honda/products/robotics.html Honda Robotics] is a robotics division of Honda based in Japan. They are known for developing the advanced humanoid robot [[ASIMO]]. {{infobox company | name = Honda Robotics | country = Japan | website_link = https://global.honda/products/robotics.html | robots = [[ASIMO]] }} [[Category:Companies]] a9af9fc3a6032570aa5052e7d912da16829198cb 0 239 1014 2024-05-14T17:46:15Z Vrtnis 21 Created page with "ASIMO is a humanoid robot developed by [[Honda Robotics]], a division of the Japanese automotive manufacturer Honda. ASIMO is one of the most advanced humanoid robots and is k..." wikitext text/x-wiki ASIMO is a humanoid robot developed by [[Honda Robotics]], a division of the Japanese automotive manufacturer Honda. ASIMO is one of the most advanced humanoid robots and is known for its remarkable walking and running capabilities. {{infobox robot | name = ASIMO | organization = [[Honda Robotics]] | height = 130 cm (4 ft 3 in) | weight = 54 kg (119 lb) | video_link = https://www.youtube.com/watch?v=Q3C5sc8EIVI | cost = Unknown }} == Development == Honda initiated the development of ASIMO in the 1980s, with the goal of creating a humanoid robot that could assist people with everyday tasks. ASIMO, an acronym for Advanced Step in Innovative Mobility, was first introduced in 2000 and has undergone several upgrades since then. == Design == ASIMO stands at a height of 130 cm and weighs approximately 54 kilograms. Its design focuses on mobility, allowing it to walk, run, climb stairs, and perform a variety of movements that mimic human actions. It is equipped with advanced sensors and actuators to navigate and interact with its environment safely. == Features == ASIMO features advanced recognition technologies, including the ability to recognize faces, voices, and gestures. It can navigate complex environments, avoid obstacles, and perform tasks such as carrying objects, opening doors, and pushing carts. ASIMO's user-friendly interface and human-like movements make it suitable for a range of applications, from public demonstrations to potential domestic assistance. == Impact == ASIMO has been influential in the field of robotics, demonstrating the potential for humanoid robots to assist in daily activities and enhance human-robot interaction. It has also played a significant role in Honda's research and development of robotics technologies, contributing to advancements in mobility, AI, and automation. == References == [https://global.honda/products/robotics.html Honda Robotics official website] [https://www.youtube.com/watch?v=Q3C5sc8EIVI Presentation of ASIMO by Honda] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Honda Robotics]] d13a632d58f80b7b05870c1186e4578a8a2d3071 1015 1014 2024-05-14T17:47:47Z Vrtnis 21 wikitext text/x-wiki ASIMO is a humanoid robot developed by [[Honda Robotics]], a division of the Japanese automotive manufacturer Honda. ASIMO is one of the more advanced humanoid robots and is known for its remarkable walking and running capabilities. {{infobox robot | name = ASIMO | organization = [[Honda Robotics]] | height = 130 cm (4 ft 3 in) | weight = 54 kg (119 lb) | video_link = https://www.youtube.com/watch?v=OkA4foR7bbk | cost = Unknown }} == Development == Honda initiated the development of ASIMO in the 1980s, with the goal of creating a humanoid robot that could assist people with everyday tasks. ASIMO, an acronym for Advanced Step in Innovative Mobility, was first introduced in 2000 and has undergone several upgrades since then. == Design == ASIMO stands at a height of 130 cm and weighs approximately 54 kilograms. Its design focuses on mobility, allowing it to walk, run, climb stairs, and perform a variety of movements that mimic human actions. It is equipped with advanced sensors and actuators to navigate and interact with its environment safely. == Features == ASIMO features advanced recognition technologies, including the ability to recognize faces, voices, and gestures. It can navigate complex environments, avoid obstacles, and perform tasks such as carrying objects, opening doors, and pushing carts. ASIMO's user-friendly interface and human-like movements make it suitable for a range of applications, from public demonstrations to potential domestic assistance. == Impact == ASIMO has been influential in the field of robotics, demonstrating the potential for humanoid robots to assist in daily activities and enhance human-robot interaction. It has also played a significant role in Honda's research and development of robotics technologies, contributing to advancements in mobility, AI, and automation. == References == [https://global.honda/products/robotics.html Honda Robotics official website] [https://www.youtube.com/watch?v=Q3C5sc8EIVI Presentation of ASIMO by Honda] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Honda Robotics]] b5c34c69e72946434e5ce78bc00e829f0cf59a9c 1016 1015 2024-05-14T17:48:07Z Vrtnis 21 wikitext text/x-wiki ASIMO is a humanoid robot developed by [[Honda Robotics]], a division of the Japanese automotive manufacturer Honda. ASIMO is one of the more advanced humanoid robots and is known for its remarkable walking and running capabilities. {{infobox robot | name = ASIMO | organization = [[Honda Robotics]] | height = 130 cm (4 ft 3 in) | weight = 54 kg (119 lb) | video_link = https://www.youtube.com/watch?v=OkA4foR7bbk | cost = Unknown }} == Development == Honda initiated the development of ASIMO in the 1980s, with the goal of creating a humanoid robot that could assist people with everyday tasks. ASIMO, an acronym for Advanced Step in Innovative Mobility, was first introduced in 2000 and has undergone several upgrades since then. == Design == ASIMO stands at a height of 130 cm and weighs approximately 54 kilograms. Its design focuses on mobility, allowing it to walk, run, climb stairs, and perform a variety of movements that mimic human actions. It is equipped with advanced sensors and actuators to navigate and interact with its environment safely. == Features == ASIMO features advanced recognition technologies, including the ability to recognize faces, voices, and gestures. It can navigate complex environments, avoid obstacles, and perform tasks such as carrying objects, opening doors, and pushing carts. ASIMO's user-friendly interface and human-like movements make it suitable for a range of applications, from public demonstrations to potential domestic assistance. == Impact == ASIMO has been influential in the field of robotics, demonstrating the potential for humanoid robots to assist in daily activities and enhance human-robot interaction. It has also played a significant role in Honda's research and development of robotics technologies, contributing to advancements in mobility, AI, and automation. == References == [https://global.honda/products/robotics.html Honda Robotics official website] [https://www.youtube.com/watch?v=Q3C5sc8EIVI Presentation of ASIMO by Honda] [[Category:Robots]] 99403ef69c95c304c11fc111e41cd8590e6254aa 0 1 1017 996 2024-05-14T17:49:00Z Vrtnis 21 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |} 7817794212da50561023a9bba635198d3259a7b9 1023 1017 2024-05-15T03:06:10Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} e43add7d64cc87dc9bb492bcc764d784001b8c9f 1026 1023 2024-05-15T19:18:10Z Ben 2 Move list of humanoid robots up higher wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} b48df7d338fb15eac12f6e336f39060ec79e6f2d 1034 1026 2024-05-15T19:35:51Z Vrtnis 21 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Reinforcement Learning]] | Resources related to understanding reinforcement learning |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 2c926d658882abfef90055bc3612cd3603eb98ec 1037 1034 2024-05-15T22:43:48Z Budzianowski 19 /* Resources */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[MyActuator X-Series]] | MIT Cheetah-like quasi-direct drive actuator, with planetary gears |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 751b964bb4b17ebeee38374b13adaf0fc24b3446 1044 1037 2024-05-16T16:19:45Z Ben 2 /* List of Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] ==== Resources ==== {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 00a93ee6944d43acd5dea68aa2dc0f9071e18547 1045 1044 2024-05-16T17:37:59Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 32c68423867181eb55e174ce61c5eaa627dfd14f 0 235 1018 1001 2024-05-14T20:18:03Z Vedant 24 wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === * How do we connect I2C chip to the LCD screen? === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery 0617fff5c391a84cb45cb113c9559124d26ddf85 1025 1018 2024-05-15T05:22:24Z Ben 2 wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === * How do we connect I2C chip to the LCD screen? === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery === Questions === # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? 9a518a08c5e61fcf2c20fd2ccc999b8210c594d1 6 240 1019 2024-05-14T21:11:17Z Ben 2 wikitext text/x-wiki Unitree g1 9157f7dba8762187f74db4e9a5a3a4c9ec00a661 10 28 1020 182 2024-05-14T21:14:26Z Ben 2 wikitext text/x-wiki {{infobox | name = {{{name}}} | key1 = Name | value1 = {{{name}}} | key2 = Organization | value2 = {{{organization|}}} | key3 = Video | value3 = {{#if: {{{video_link|}}} | [{{{video_link}}} Video] }} | key4 = Cost | value4 = {{{cost|}}} | key5 = Height | value5 = {{{height|}}} | key6 = Weight | value6 = {{{weight|}}} | key7 = Speed | value7 = {{{speed|}}} | key8 = Lift Force | value8 = {{{lift_force|}}} | key9 = Battery Life | value9 = {{{battery_life|}}} | key10 = Battery Capacity | value10 = {{{battery_capacity|}}} | key11 = Purchase | value11 = {{#if: {{{purchase_link|}}} | [{{{purchase_link}}} Link] }} | key12 = Number Made | value12 = {{{number_made|}}} | key13 = DoF | value13 = {{{dof|}}} | key14 = Status | value14 = {{{status|}}} | key15 = Notes | value15 = {{{notes|}}} }} ae8d9b2da5170a553c8ba59d56d16d988c18821e 0 233 1021 997 2024-05-14T21:16:43Z Ben 2 wikitext text/x-wiki [[File:Unitree g1.png|thumb]] The G1 humanoid robot is an upcoming humanoid robot from [[Unitree]]. {{infobox robot | name = G1 | organization = [[Unitree]] | video_link = https://mp.weixin.qq.com/s/RGNVRazZqDn3y_Ijemc5Kw | cost = 16000 USD | height = 127 cm | weight = 35 kg | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = 23 | status = Preorders }} {{infobox robot | name = G1 Edu Standard | organization = [[Unitree]] | cost = 31900 USD | notes = Improved torque, warranty }} {{infobox robot | name = G1 Edu Plus | organization = [[Unitree]] | cost = 34900 USD | notes = Docking station }} {{infobox robot | name = G1 Edu Smart | organization = [[Unitree]] | cost = 43900 USD | notes = 3 waist DoFs instead of 1, more arm DoFs }} {{infobox robot | name = G1 Edu Ultimate | organization = [[Unitree]] | cost = 53900 USD | notes = Comes with force-controlled 3-finger dexterous hands }} 11b1e60701dbf141ea569693269ca05438fb79c1 1022 1021 2024-05-14T21:20:19Z Ben 2 wikitext text/x-wiki [[File:Unitree g1.png|thumb]] The G1 humanoid robot is an upcoming humanoid robot from [[Unitree]]. {{infobox robot | name = G1 | organization = [[Unitree]] | video_link = https://mp.weixin.qq.com/s/RGNVRazZqDn3y_Ijemc5Kw | cost = 16000 USD | height = 127 cm | weight = 35 kg | speed = | lift_force = | battery_life = | battery_capacity = 9000 mAh | purchase_link = | number_made = | dof = 23 | status = Preorders }} {{infobox robot | name = G1 Edu Standard | organization = [[Unitree]] | cost = 31900 USD | notes = Improved torque, warranty }} {{infobox robot | name = G1 Edu Plus | organization = [[Unitree]] | cost = 34900 USD | notes = Docking station }} {{infobox robot | name = G1 Edu Smart | organization = [[Unitree]] | cost = 43900 USD | notes = 3 waist DoFs instead of 1, more arm DoFs }} {{infobox robot | name = G1 Edu Ultimate | organization = [[Unitree]] | cost = 53900 USD | notes = Comes with force-controlled 3-finger dexterous hands }} 303bd7bc226ac0180d7db51ff129b963fff0a01a 0 241 1024 2024-05-15T03:06:40Z Modeless 7 Created page with "A humanoid built at the University of Tehran. https://surenahumanoid.com/" wikitext text/x-wiki A humanoid built at the University of Tehran. https://surenahumanoid.com/ 61aa82a1b4c689d70dffa1f49952dc7b06cb0f8a 0 242 1027 2024-05-15T19:22:06Z Vrtnis 21 Created page with "[https://www.softbankrobotics.com/ SoftBank Robotics] is a robotics division of SoftBank Group, based in Japan. They are known for developing advanced humanoid robots such as..." wikitext text/x-wiki [https://www.softbankrobotics.com/ SoftBank Robotics] is a robotics division of SoftBank Group, based in Japan. They are known for developing advanced humanoid robots such as [[Pepper]] and [[NAO]]. {{infobox company | name = SoftBank Robotics | country = Japan | website_link = https://www.softbankrobotics.com/ | robots = [[Pepper]], [[NAO]] }} [[Category:Companies]] 74d615743cf5ea836b0ef26ac27fae0687e453b3 1032 1027 2024-05-15T19:34:20Z Vrtnis 21 Vrtnis moved page [[Softbank Robotics]] to [[SoftBank Robotics]]: Correct title capitalization wikitext text/x-wiki [https://www.softbankrobotics.com/ SoftBank Robotics] is a robotics division of SoftBank Group, based in Japan. They are known for developing advanced humanoid robots such as [[Pepper]] and [[NAO]]. {{infobox company | name = SoftBank Robotics | country = Japan | website_link = https://www.softbankrobotics.com/ | robots = [[Pepper]], [[NAO]] }} [[Category:Companies]] 74d615743cf5ea836b0ef26ac27fae0687e453b3 0 243 1028 2024-05-15T19:25:00Z Vrtnis 21 Created page with "Pepper is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. Pepper is designed to interact with humans and is used in various customer service..." wikitext text/x-wiki Pepper is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. Pepper is designed to interact with humans and is used in various customer service and retail environments. {{infobox robot | name = Pepper | organization = [[Softbank Robotics]] | height = 121 cm (4 ft) | weight = 28 kg (62 lbs) | video_link = https://www.youtube.com/watch?v=2GhUd0OJdJw | cost = Approximately $1,800 }} Pepper was introduced by SoftBank Robotics in June 2014. It is designed to understand and respond to human emotions, making it suitable for roles in customer service, retail, and healthcare. == References == [https://www.softbankrobotics.com/ SoftBank Robotics official website] [https://www.youtube.com/watch?v=2GhUd0OJdJw Presentation of Pepper by SoftBank Robotics] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:SoftBank Robotics]] a28a621fc3e7afb36cc3d20c6fdb5b733e8b49b8 1029 1028 2024-05-15T19:26:11Z Vrtnis 21 wikitext text/x-wiki Pepper is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. Pepper is designed to interact with humans and is used in various customer service and retail environments. {{infobox robot | name = Pepper | organization = [[Softbank Robotics]] | height = 121 cm (4 ft) | weight = 28 kg (62 lbs) | video_link = https://www.youtube.com/watch?v=kr05reBxVRs | cost = Approximately $1,800 }} Pepper was introduced by SoftBank Robotics in June 2014. It is designed to understand and respond to human emotions, making it suitable for roles in customer service, retail, and healthcare. == References == [https://www.softbankrobotics.com/ SoftBank Robotics official website] [https://www.youtube.com/watch?v=2GhUd0OJdJw Presentation of Pepper by SoftBank Robotics] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:SoftBank Robotics]] 9e0192c414352ae95ba759d3fc4088798349a1b1 1030 1029 2024-05-15T19:26:31Z Vrtnis 21 /* References */ wikitext text/x-wiki Pepper is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. Pepper is designed to interact with humans and is used in various customer service and retail environments. {{infobox robot | name = Pepper | organization = [[Softbank Robotics]] | height = 121 cm (4 ft) | weight = 28 kg (62 lbs) | video_link = https://www.youtube.com/watch?v=kr05reBxVRs | cost = Approximately $1,800 }} Pepper was introduced by SoftBank Robotics in June 2014. It is designed to understand and respond to human emotions, making it suitable for roles in customer service, retail, and healthcare. == References == [https://www.softbankrobotics.com/ SoftBank Robotics official website] [https://www.youtube.com/watch?v=2GhUd0OJdJw Presentation of Pepper by SoftBank Robotics] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:SoftBank Robotics]] 726c2f29bafc2eea3f45c73e4932e7c23476cfea 0 244 1031 2024-05-15T19:29:50Z Vrtnis 21 Created page with "NAO is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. NAO is widely used in education, research, and healthcare for its advanced interactiv..." wikitext text/x-wiki NAO is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. NAO is widely used in education, research, and healthcare for its advanced interactive capabilities. {{infobox robot | name = NAO | organization = [[Softbank Robotics]] | height = 58 cm (1 ft 11 in) | weight = 5.4 kg (11.9 lbs) | video_link = https://www.youtube.com/watch?v=nNbj2G3GmAo | cost = Approximately $8,000 }} NAO was first introduced in 2006 by Aldebaran Robotics, which was later acquired by SoftBank Robotics. NAO has undergone several upgrades, becoming one of the most popular robots used for educational and research purposes. == References == [https://www.softbankrobotics.com/ SoftBank Robotics official website] [https://www.youtube.com/watch?v=nNbj2G3GmAo Presentation of NAO by SoftBank Robotics] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:SoftBank Robotics]] 9e28814a20e1f1e0f11d3a085b90f14964143c74 0 245 1033 2024-05-15T19:34:20Z Vrtnis 21 Vrtnis moved page [[Softbank Robotics]] to [[SoftBank Robotics]]: Correct title capitalization wikitext text/x-wiki #REDIRECT [[SoftBank Robotics]] 6519847bc798ec52ddbf466a6d15bfd308315c52 0 32 1035 341 2024-05-15T20:14:33Z Vrtnis 21 /* Isaac Sim */ wikitext text/x-wiki Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with example [[applications]]. Typically you need a simulator, training framework and machine learning method to train end to end behaviors. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. For a much more comprehensive overview see [https://simulately.wiki/docs/ Simulately]. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===Mujoco=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ===Bullet=== Bullet is a physics engine supporting real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning ==Simulators== ===[[Isaac Sim]]=== Isaac Sim is NVIDIA’s simulation platform for robotics development. It’s part of their Isaac Robotics platform and uses advanced graphics and AI to create realistic simulations. ==== Isaac Sim Features ==== * '''Advanced Physics Simulation''': Includes PhysX and Flex for detailed simulations of physical interactions like rigid bodies, soft bodies, and fluids. * '''Photorealistic Rendering''': Uses NVIDIA RTX technology to make environments and objects look incredibly realistic, which is great for tasks that need vision-based learning. * '''Scalability''': Can simulate multiple robots and environments at the same time, thanks to GPU acceleration, making it handle complex simulations efficiently. * '''Interoperability''': Works with machine learning frameworks like TensorFlow and PyTorch and supports ROS, so you can easily move from simulation to real-world deployment. * '''Customizable Environments''': Lets you create and customize simulation environments, including importing 3D models and designing different terrains. * '''Real-Time Feedback''': Provides real-time monitoring and analytics, giving you insights on how tasks are performing and resource usage. ==== Isaac Sim Applications ==== * '''Robotics Research''': Used in academia and industry to develop and test new algorithms for robot perception, control, and planning. * '''Autonomous Navigation''': Helps simulate and test navigation algorithms for mobile robots and drones, improving path planning and obstacle avoidance. * '''Manipulation Tasks''': Supports developing robotic skills like object grasping and assembly tasks, making robots more dexterous and precise. * '''Industrial Automation''': Helps companies design and validate automation solutions for manufacturing and logistics, boosting efficiency and cutting down on downtime. * '''Education and Training''': A great educational tool that offers hands-on experience in robotics and AI without the risks and costs of physical experiments. === Isaac Sim Integration with Isaac Gym === Isaac Sim works alongside Isaac Gym, NVIDIA’s tool for large-scale training with reinforcement learning. While Isaac Sim focuses on detailed simulations, Isaac Gym is great for efficient training. Together, they offer a comprehensive solution for developing and improving robotics applications. ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== =Training frameworks= Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 4c7025f8eee4e7dc648d770d3158af6d142cfa15 4 17 1036 55 2024-05-15T22:15:45Z 204.15.110.167 0 wikitext text/x-wiki This is Gregory Stewart from Wyoming. We'd like to offer your business a loan to kick off the new year, to use for whatever you need. We're reaching out to a few local companies and I just wanted to see if we can help at all. Please take a look the details I put on our page here - https://cutt.ly/lwHyBuO7 All the best, Gregory Stewart - Owner Fast Money Locator, LLC 79fa7f878be7f3098510bab1aa64f111fd6de211 0 16 1038 930 2024-05-16T00:24:34Z Ben 2 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. == Reserving a GPU == Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. === Useful Commands === Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] 34dec652748a6a923c07d8d18b7884991feac81a 1039 1038 2024-05-16T00:25:53Z Ben 2 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. === Useful Commands === Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] 8685d8c0b9d53d80ea92ef568b2d36f92ed4edaf 1040 1039 2024-05-16T00:26:13Z Ben 2 wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. ==== Useful Commands ==== Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] 05b83ad5f11753866813aef9d48b28510dedf218 0 34 1041 759 2024-05-16T06:05:08Z Vrtnis 21 /* Training algorithms */ wikitext text/x-wiki == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] (also see slides on Actor Critic methods at [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf] Stanford CS224R) * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] [[Category: Software]] a05cd43f4cf069c654ffb9ff62c2e3c06d46b99d 1042 1041 2024-05-16T06:21:21Z Vrtnis 21 /* Training algorithms */ wikitext text/x-wiki == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] (also see slides on Actor Critic methods at [1]) * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == References == * [1] [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] [[Category: Software]] dfbb941d25dc755b1e4d25df186c10e7ff36c9ae 1043 1042 2024-05-16T06:22:52Z Vrtnis 21 wikitext text/x-wiki == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] 81b6d22b02fc91d39136eccf3357bcf77d4533b0 1050 1043 2024-05-16T19:54:18Z Vrtnis 21 wikitext text/x-wiki == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions, and using this feedback to improve its performance over time. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] 878f1cc05b6d90e89c211e5a972dd1c1098014ab 1051 1050 2024-05-16T19:56:42Z Vrtnis 21 wikitext text/x-wiki == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] a7bf02ec47c062772e3d6b206665bc477e1e6381 1052 1051 2024-05-16T19:57:42Z Vrtnis 21 /* Adding Isaac related details to Reinforcement Learning (RL) */ wikitext text/x-wiki == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == RL in Robotics == === Practical Applications of RL === ==== Task Automation ==== * Robots can be trained to perform repetitive or dangerous tasks autonomously, such as assembly line work, welding, or hazardous material handling. * RL enables robots to adapt to new tasks without extensive reprogramming, making them versatile for various industrial applications. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] 3d65c0fdedca6b5a4db552e3b149e27f87ac7cf1 1053 1052 2024-05-16T20:01:06Z Vrtnis 21 wikitext text/x-wiki == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == RL in Robotics == === Practical Applications of RL === ==== Task Automation ==== * Robots can be trained to perform repetitive or dangerous tasks autonomously, such as assembly line work, welding, or hazardous material handling. * RL enables robots to adapt to new tasks without extensive reprogramming, making them versatile for various industrial applications. ==== Navigation and Manipulation ==== * RL is used to train robots for navigating complex environments and manipulating objects with precision, which is crucial for tasks like warehouse logistics, domestic chores, and medical surgeries. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] 1eed9af76f858a24b5378576677f6dd810f39437 1054 1053 2024-05-16T20:02:10Z Vrtnis 21 wikitext text/x-wiki == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == RL in Robotics == === Practical Applications of RL === ==== Task Automation ==== * Robots can be trained to perform repetitive or dangerous tasks autonomously, such as assembly line work, welding, or hazardous material handling. * RL enables robots to adapt to new tasks without extensive reprogramming, making them versatile for various industrial applications. ==== Navigation and Manipulation ==== * RL is used to train robots for navigating complex environments and manipulating objects with precision, which is crucial for tasks like warehouse logistics, domestic chores, and medical surgeries. === Simulation Environments === ==== Isaac Sim ==== * Isaac Sim provides a highly realistic and interactive environment where robots can be trained safely and efficiently. * The simulated environment includes physics, sensors, and other elements that mimic real-world conditions, enabling the transfer of learned behaviors to physical robots. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] 3c56c8092b10d48a7e53c8e9d9209b562dfeb800 1055 1054 2024-05-16T20:03:33Z Vrtnis 21 /* Add work in progress tag at top */ wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == RL in Robotics == === Practical Applications of RL === ==== Task Automation ==== * Robots can be trained to perform repetitive or dangerous tasks autonomously, such as assembly line work, welding, or hazardous material handling. * RL enables robots to adapt to new tasks without extensive reprogramming, making them versatile for various industrial applications. ==== Navigation and Manipulation ==== * RL is used to train robots for navigating complex environments and manipulating objects with precision, which is crucial for tasks like warehouse logistics, domestic chores, and medical surgeries. === Simulation Environments === ==== Isaac Sim ==== * Isaac Sim provides a highly realistic and interactive environment where robots can be trained safely and efficiently. * The simulated environment includes physics, sensors, and other elements that mimic real-world conditions, enabling the transfer of learned behaviors to physical robots. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] 11e85e3de6bfb760b73916e6ebbb12a15713e207 1056 1055 2024-05-16T20:06:41Z Vrtnis 21 /* Add MuJoCo*/ wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == RL in Robotics == === Practical Applications of RL === ==== Task Automation ==== * Robots can be trained to perform repetitive or dangerous tasks autonomously, such as assembly line work, welding, or hazardous material handling. * RL enables robots to adapt to new tasks without extensive reprogramming, making them versatile for various industrial applications. ==== Navigation and Manipulation ==== * RL is used to train robots for navigating complex environments and manipulating objects with precision, which is crucial for tasks like warehouse logistics, domestic chores, and medical surgeries. === Simulation Environments === ==== Isaac Sim ==== * Isaac Sim provides a highly realistic and interactive environment where robots can be trained safely and efficiently. * The simulated environment includes physics, sensors, and other elements that mimic real-world conditions, enabling the transfer of learned behaviors to physical robots. ==== MuJoCo ==== * MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research and development in robotics, machine learning, and biomechanics. * It offers fast and accurate simulations, which are essential for training RL agents in tasks involving complex dynamics and contact-rich interactions. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] 7376f8bb8c449a6908ca7501563ac19e8e908671 0 139 1046 999 2024-05-16T18:14:23Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Onboarding === * Watch out for an email from Gusto (our HR software), with an official offer letter and instructions on how to onboard you into our system. * Once you accept, Ben will add you to the system, after which you will have to enter your bank account information in order to be paid. * Gusto will send you an email the day before your expected start date with additional on-boarding tasks. If you are waiting on an email which you haven't received yet, odds are it is just not needed yet === Pre-Internship Checklist === * Create a wiki account and mark yourself as an employee (you can use [[User:Ben]] as a template). You'll use your account as the main way to keep track of what you've done over the course of the internship. * Contribute an article about something you find interesting. See the [[Contributing]] guide. === What To Bring === * Bring living things (clothing, toothbrush, etc.) ** We will have beds, bedsheets, towels * For your first day, you will need documents for your I9 authorizing you to work with us. The easiest is to just bring your passport or passport card. Alternatively, you'll need your driver's license or a federal photo ID, AND your social security card or birth certificate. === Arrival === * We have a company standup every day at 8:45 AM * Arrive anytime prior to your start date === Additional Notes === * For travel expenses, please purchase your own flight and keep your receipts so that we can reimburse you later. [[Category:K-Scale]] e41b7f10c0fdffccea25597079888e21ea6bd098 1047 1046 2024-05-16T18:15:52Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Onboarding === * Watch out for an email from Gusto (our HR software), with an official offer letter and instructions on how to onboard you into our system. * Once you accept, Ben will add you to the system, after which you will have to enter your bank account information in order to be paid. * Gusto will send you an email the day before your expected start date with additional on-boarding tasks. If you are waiting on an email which you haven't received yet, odds are it is just not needed yet === Pre-Internship Checklist === * Create a wiki account and mark yourself as an employee (you can use [[User:Ben]] as a template). You'll use your account as the main way to keep track of what you've done over the course of the internship. * Contribute an article about something you find interesting. See the [[Contributing]] guide. * For your first day, you will need documents for your I9 authorizing you to work with us. The easiest is to just bring your passport or passport card. Alternatively, you'll need your driver's license or a federal photo ID, AND your social security card or birth certificate. === What To Bring === * Bring living things (clothing, toothbrush, etc.) * We will have beds, bedsheets, towels, toothpaste, and shampoo for you. Other miscellaneous toiletries will also be taken care of. === Arrival === * We have a company standup every day at 8:45 AM. * Arrive anytime prior to your start date. If you want to come early that is fine as well. === Expenses === * For travel expenses, please purchase your own flight and keep your receipts so that we can reimburse you later. [[Category:K-Scale]] 587b281ded6114dc934b578775bc7a58fc7251ed 1048 1047 2024-05-16T18:22:09Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. We've moved our onboarding checklist to our [https://wiki.kscale.dev/w/Intern_Onboarding internal wiki]. Please reach out to Ben to get your credentials. [[Category:K-Scale]] 9a3c1bb090edc0d701f79c074d99de83be20ec57 1049 1048 2024-05-16T18:23:21Z Ben 2 wikitext text/x-wiki Congratulations on your internship at K-Scale Labs! We are excited for you to join us. === Pre-Internship Checklist === * Create a wiki account and mark yourself as an employee (you can use [[User:Ben]] as a template). You'll use your account as the main way to keep track of what you've done over the course of the internship. * Contribute an article about something you find interesting. See the [[Contributing]] guide. * Once you get your credentials, follow the onboarding checklist to our [https://wiki.kscale.dev/w/Intern_Onboarding internal wiki]. Please reach out to Ben if you have any questions. [[Category:K-Scale]] 37a0733da7a987351e3d80ad4a88a2eaa25d8145 0 193 1057 941 2024-05-16T20:08:53Z Vrtnis 21 /* Added wiki-link to RL page */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the [https://humanoids.wiki/w/MIT_Cheetah MIT Cheetah] actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost.If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== [[Reinforcement Learning]] Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the MuJoCo physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although MuJoCo can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. More resources are available at [https://humanoids.wiki/w/Learning_algorithms Learning Algorithms] == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to an open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs. Check out [https://humanoids.wiki/w/Stompy Stompy!] === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 0c8ba42aeac159bde9eb89e1f302dc3cf63fecf3 0 32 1058 1035 2024-05-16T20:13:27Z Vrtnis 21 /* Mujoco */ wikitext text/x-wiki Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with example [[applications]]. Typically you need a simulator, training framework and machine learning method to train end to end behaviors. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. For a much more comprehensive overview see [https://simulately.wiki/docs/ Simulately]. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===MuJoCo=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ===Bullet=== Bullet is a physics engine supporting real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning ==Simulators== ===[[Isaac Sim]]=== Isaac Sim is NVIDIA’s simulation platform for robotics development. It’s part of their Isaac Robotics platform and uses advanced graphics and AI to create realistic simulations. ==== Isaac Sim Features ==== * '''Advanced Physics Simulation''': Includes PhysX and Flex for detailed simulations of physical interactions like rigid bodies, soft bodies, and fluids. * '''Photorealistic Rendering''': Uses NVIDIA RTX technology to make environments and objects look incredibly realistic, which is great for tasks that need vision-based learning. * '''Scalability''': Can simulate multiple robots and environments at the same time, thanks to GPU acceleration, making it handle complex simulations efficiently. * '''Interoperability''': Works with machine learning frameworks like TensorFlow and PyTorch and supports ROS, so you can easily move from simulation to real-world deployment. * '''Customizable Environments''': Lets you create and customize simulation environments, including importing 3D models and designing different terrains. * '''Real-Time Feedback''': Provides real-time monitoring and analytics, giving you insights on how tasks are performing and resource usage. ==== Isaac Sim Applications ==== * '''Robotics Research''': Used in academia and industry to develop and test new algorithms for robot perception, control, and planning. * '''Autonomous Navigation''': Helps simulate and test navigation algorithms for mobile robots and drones, improving path planning and obstacle avoidance. * '''Manipulation Tasks''': Supports developing robotic skills like object grasping and assembly tasks, making robots more dexterous and precise. * '''Industrial Automation''': Helps companies design and validate automation solutions for manufacturing and logistics, boosting efficiency and cutting down on downtime. * '''Education and Training''': A great educational tool that offers hands-on experience in robotics and AI without the risks and costs of physical experiments. === Isaac Sim Integration with Isaac Gym === Isaac Sim works alongside Isaac Gym, NVIDIA’s tool for large-scale training with reinforcement learning. While Isaac Sim focuses on detailed simulations, Isaac Gym is great for efficient training. Together, they offer a comprehensive solution for developing and improving robotics applications. ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== =Training frameworks= Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 6116a27ba4c142dcbac4fe72994cab9c50891ea8 0 246 1059 2024-05-16T20:15:29Z Ben 2 Created page with "Mirasol3B is an autoregressive multimodal model for time-aligned video and audio. {{infobox paper | name = Mirasol3B | full_name = Mirasol3B: A Multimodal Autoregressive Mode..." wikitext text/x-wiki Mirasol3B is an autoregressive multimodal model for time-aligned video and audio. {{infobox paper | name = Mirasol3B | full_name = Mirasol3B: A Multimodal Autoregressive Model for Time-Aligned and Contextual Modalities | arxiv_link = https://arxiv.org/abs/2311.05698 | project_page = https://research.google/blog/scaling-multimodal-understanding-to-long-videos/ | twitter_link = https://twitter.com/GoogleAI/status/1724553024088191211 | date = February 2024 | authors = AJ Piergiovanni, Isaac Noble, Dahun Kim, Michael Ryoo, Victor Gomes, Anelia Angelova }} [[Category: Papers]] 062a79b195d8f90b4bb441a3b59b76c2fe35b2bb 0 1 1060 1045 2024-05-16T20:17:02Z Ben 2 /* Getting Started */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 9f9671e2fab483fe719415d49f496f9b20bc2dce 1078 1060 2024-05-19T16:55:57Z 46.193.2.72 0 /* Getting Started */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [[Advanced Robot Dynamics]] | High-quality open-source course from CMU |- | [[Optimal Control]] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} adfab4feab57685f1852ec89ed6a5bdef886f95e 1079 1078 2024-05-19T16:57:55Z 46.193.2.72 0 /* Getting Started */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 07826297e96411b49b2071e9b33f9650fea8b810 0 247 1061 2024-05-16T20:20:15Z Vrtnis 21 Created page with "== MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. T..." wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. 2424281c488950aad97c269811beb8f5e9f45fc9 1063 1061 2024-05-16T20:28:12Z Vrtnis 21 /* Add suggestions and tips */ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. 3a8627eaea22adf82d3545041ef51b40f3e28175 1066 1063 2024-05-17T04:36:02Z Vrtnis 21 /* Added Tips and Suggestions */ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. 50b83646aa82e6a991c697d695f0a3409283ba2f 1067 1066 2024-05-17T04:36:39Z Vrtnis 21 /*Added Tips and Suggestions */ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. 90fcf5ead620a4c480c2ddd481ff58ef17804251 1068 1067 2024-05-17T04:37:59Z Vrtnis 21 /* Added tips and Suggestions */ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. d3ceb006286874080787404c1d363cc0fe71b10c 1072 1068 2024-05-19T15:52:30Z Vrtnis 21 /* Add MuJoCo Menagerie*/ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. 50204989921edabd6c890401f3ce621e5b32f52c 1073 1072 2024-05-19T16:11:00Z Vrtnis 21 /* Add MuJoCo Menagerie Getting Started*/ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. === Getting Started === Download models and example files from the official MuJoCo website or repository. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. 1f3e9cda63f1f89bc4c7818d37167c45a87a1d56 1074 1073 2024-05-19T16:25:21Z Vrtnis 21 /* Add code samples for Menagerie*/ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. === Getting Started === Download models and example files from the official MuJoCo website or repository. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. === Getting Started === To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. d6366316d0a4387729bb01946a9e9e9f1e506daf 1075 1074 2024-05-19T16:25:43Z Vrtnis 21 wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. === Getting Started === Download models and example files from the official MuJoCo website or repository. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. === Getting Started === To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> 8138bba2632d5fd6de12b7825497066c15142fba 1077 1075 2024-05-19T16:55:42Z Vrtnis 21 /* MuJoCo Menagerie */ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. [[File:Mujoco_menagerie_github.png|alt=Source: MuJoCo Menagerie GitHub|none|300px|Source: MuJoCo Menagerie GitHub]] === Getting Started === Download models and example files from the official MuJoCo website or repository. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. === Getting Started === To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> 10abf6c10ee23f8a35b3c80f8ee36bc592662547 1080 1077 2024-05-19T16:58:17Z Vrtnis 21 /* Add repo link */ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. [[File:Mujoco_menagerie_github.png|alt=Source: MuJoCo Menagerie GitHub|none|300px|Source: MuJoCo Menagerie GitHub]] === Getting Started === Download models and example files from the official MuJoCo website or [https://github.com/google-deepmind/mujoco_menagerie repository]. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. === Getting Started === To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> 57bf886d7b8220229c1632c921786f5e8516f1fe 1081 1080 2024-05-19T16:58:40Z Vrtnis 21 /* Getting Started */ wikitext text/x-wiki == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. [[File:Mujoco_menagerie_github.png|alt=Source: MuJoCo Menagerie GitHub|none|300px|Source: MuJoCo Menagerie GitHub]] === Getting Started === Download models and example files from the official MuJoCo website or [https://github.com/google-deepmind/mujoco_menagerie repository]. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> d72e6c391b4620e4344bdc91690b9d403f687890 1082 1081 2024-05-19T16:59:27Z Vrtnis 21 wikitext text/x-wiki This is incomplete and a work in progress; you can help by expanding it! == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This open-source software provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. [[File:Mujoco_menagerie_github.png|alt=Source: MuJoCo Menagerie GitHub|none|300px|Source: MuJoCo Menagerie GitHub]] === Getting Started === Download models and example files from the official MuJoCo website or [https://github.com/google-deepmind/mujoco_menagerie repository]. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> f23193e77b902e550e1129ee80b67742e10d85a4 1083 1082 2024-05-19T17:01:22Z Vrtnis 21 wikitext text/x-wiki This is incomplete and a work in progress; you can help by expanding it! == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This [https://github.com/google-deepmind/mujoco open-source software] provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. [[File:Mujoco_menagerie_github.png|alt=Source: MuJoCo Menagerie GitHub|none|300px|Source: MuJoCo Menagerie GitHub]] === Getting Started === Download models and example files from the official MuJoCo website or [https://github.com/google-deepmind/mujoco_menagerie repository]. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> 6312f98ee95aa96ebef1e6d67ebb5884e86b6bfa 1096 1083 2024-05-19T18:52:04Z Vrtnis 21 /*Add WASM Ref*/ wikitext text/x-wiki This is incomplete and a work in progress; you can help by expanding it! == MuJoCo == MuJoCo, short for Multi-Joint dynamics with Contact, is a physics engine designed for research and development in robotics, machine learning, and biomechanics. This [https://github.com/google-deepmind/mujoco open-source software] provides accurate and efficient simulation of complex physical systems, making it highly regarded among researchers and engineers. === History === MuJoCo was developed at the University of Washington and released in 2012. Initially created to aid in the study of motor control in humans and robots, it quickly gained popularity for its ability to simulate complex physical interactions with high precision and efficiency. In 2021, DeepMind, a subsidiary of Alphabet Inc., acquired MuJoCo and made it freely available to the public. There is also a community built MuJoCo WASM port that loads and runs MuJoCo 2.3.1 Models using JavaScript and WebAssembly<ref>https://github.com/zalo/mujoco_wasm</ref>. === Tips and Suggestions === When creating MuJoCo XML files, it can be helpful to start with a full robot model and then manually copy and paste sections to create single or dual-arm setups. This approach saves time and ensures consistency across different models. Additionally, MuJoCo supports defining keyframes for specific poses directly in the XML, which can streamline the setup process for various robot configurations. If you encounter issues with simulation stability, e.g. warnings about NaN values, consider fine-tuning the hyperparameters, including control gains and regularization terms, to achieve more stable and reliable simulations. == MuJoCo Menagerie == The MuJoCo Menagerie is a collection of pre-built models and simulation environments showcasing the capabilities of the MuJoCo physics engine. This resource is valuable for researchers, engineers, and enthusiasts exploring robotics, biomechanics, and machine learning applications. [[File:Mujoco_menagerie_github.png|alt=Source: MuJoCo Menagerie GitHub|none|300px|Source: MuJoCo Menagerie GitHub]] === Getting Started === Download models and example files from the official MuJoCo website or [https://github.com/google-deepmind/mujoco_menagerie repository]. Detailed documentation and setup instructions accompany each model. Users are encouraged to experiment, modify XML files, and share their findings with the community. By leveraging the MuJoCo Menagerie, researchers and engineers can accelerate their work, explore new possibilities, and contribute to the MuJoCo ecosystem. To install the Menagerie, simply clone the repository in the directory of your choice: <syntaxhighlight lang="bash"> git clone https://github.com/google-deepmind/mujoco_menagerie.git </syntaxhighlight> The easiest way to interactively explore a model is to load it in the simulate binary which ships with every MuJoCo distribution. Just drag and drop the <code>scene.xml</code> file into the simulate window. Alternatively, you can use the command line to launch simulate and directly pass in the path to the XML. Outside of interactive simulation, you can load a model exactly as you would with any other XML file in MuJoCo. Here’s how you do it with the C/C++ API: <syntaxhighlight lang="c"> #include <mujoco.h> mjModel* model = mj_loadXML("unitree_a1/a1.xml", nullptr, nullptr, 0); mjData* data = mj_makeData(model); mj_step(model, data); </syntaxhighlight> And here’s how you do it with Python: <syntaxhighlight lang="python"> import mujoco model = mujoco.MjModel.from_xml_path("unitree_a1/a1.xml") data = mujoco.MjData(model) mujoco.mj_step(model, data) </syntaxhighlight> 3b948aa1215c69ee75bc806b08889cb964d5e56d 0 193 1062 1057 2024-05-16T20:23:55Z Vrtnis 21 /* Add wikilink to MuJoCo (MJX) */ wikitext text/x-wiki This is a build guide for getting started experimenting with your own humanoid robot. This is incomplete and a work in progress; you can help by expanding it! == Building Your Humanoid Robot == In humanoid robotics, choosing the right components, for example, actuators and gearboxes is crucial. Folks can use planetary and cycloidal gear actuators for their precision and strength, along with Series Elastic and Quasi-Direct Drive actuators for smoother, more natural movements. Advanced designs like the [https://humanoids.wiki/w/MIT_Cheetah MIT Cheetah] actuator push the boundaries with fast, agile movements. Projects like the SPIN initiative are also key, as they make high-quality actuator technology more accessible, helping the field evolve and improve. == Actuators and Gearboxes == === Actuator Types and Design Inspirations === ==== Planetary and Cycloidal Gear Actuators ==== These actuators remain popular in the robotics community due to their high torque output and compact form factors. Planetary gears are favored for their efficiency and ability to handle high power densities, crucial for humanoid robotics. Cycloidal gears offer superior load-bearing capabilities and minimal backlash, ideal for precise motion control. MyActuator (just one option) offers a variety of planetary actuators. These actuators, while still relatively pricey, offer robust performance and are integral to the efficient functioning of the builds. Some models are: RMD X4: A lightweight and compact actuator that provides precise control and high efficiency. RMD X6: Offers a good balance of torque and speed, suitable for medium-sized applications. RMD X8: Features a more powerful motor and higher torque capacity, making it ideal for more demanding tasks. RMD X10: The most powerful actuator used, designed for high torque applications with excellent control features. ==== Series Elastic and Quasi-Direct Drive Actuators ==== Series Elastic Actuators (SEAs) are used in applications requiring safe and compliant human-robot interaction. They incorporate elastic elements, allowing for energy absorption and safer interactions. Quasi-Direct Drive Actuators provide a balance between the control fidelity of direct drives and the mechanical simplicity of geared systems, promoting natural and responsive movements. Some things to consider: The springs in SEAs are where the magic happens. Choosing the right stiffness is a balancing act between getting precise torque control and avoiding sluggish responses. Since the spring is constantly flexing, you need sensors tuned to give accurate torque measurements. Do it regularly, and you'll keep those movements smooth and predictable. You want finely-tuned control loops to make SEAs shine. A high-frequency loop can make your robot more agile in handling external forces. PID controllers are a solid starting point, or you can try out some advanced strategies. Friction can really impact your torque control, especially in gearboxes and linkages. Using low-friction components and proper lubrication will help keep everything moving smoothly. Make sure your spring is positioned directly between the actuator and the joint. If not, your robot won’t get the full benefit of force sensing, and that precision will be lost.If your robot is doing a lot of high-impact activities, the springs can wear out. Keep an eye on them to avoid breakdowns when you least expect them. SEAs thrive on real-time feedback. Ensure your software can handle data quickly, maybe using a real-time operating system or optimized signal processing. ==== MIT Cheetah Actuator ==== The MIT Cheetah actuator design is a notable example that several community members are considering emulating. Its design optimizes for rapid, dynamic movements and could potentially set a standard for agile robotic locomotion. Its designed to pack a lot of power into a lightweight, compact system. It offers excellent torque and control without being bulky, making it perfect for mobile robots that need to be quick on their feet. Also, it's energy-efficient and provides a high torque-to-weight ratio, so robots can move fast and precisely, which is essential for those tricky, agile movements. One of the coolest things about this actuator is how it manages to minimize backlash, giving you smooth, accurate control over the robot's motion. Its integrated design also means the motor and controller work together seamlessly, which keeps the system streamlined. Plus, the advanced control algorithms make it easy for the actuator to handle dynamic motions, whether it's fast acceleration or sharp turns. If you're building a robot that needs to move like a sprinter while staying super nimble, the MIT Cheetah actuator is an awesome choice. Here is the [https://fab.cba.mit.edu/classes/865.18/motion/papers/mit-cheetah-actuator.pdf MIT research paper] if you are interested in a deeper dive. === Open-Source Development and Collaboration === ==== SPIN: A Revolutionary Servo Project ==== The [https://github.com/atopile/spin-servo-drive SPIN Project] by Atopile is developing an open-source hardware project aimed at making it easier and more cost-effective to use BLDC servo motors. This project is particularly notable for its potential to democratize high-quality actuator technology, making it accessible for a broader range of developers and hobbyists. === Community Insights and Future Directions === ==== Comprehensive Actuator Comparisons ==== The humanoid robotics community actively discusses the need for a universal platform to compare and contrast the cost and performance of commercially available actuators. This could involve developing a comprehensive database or chart detailing each actuator's cost per Newton-meter, control schemes, and RPM, providing a valuable resource for both newcomers and experienced developers. Here is a [https://jakeread.pages.cba.mit.edu/actuators/ scatter plot] of actuators hosted at MIT ==== Custom Actuator Developments ==== [https://irisdynamics.com/products/orca-series Iris Dynamics electric linear actuators] suggest they can match the capabilities of human muscles, making them particularly interesting for humanoid applications. == Assembly Tips == ===== Community Forums ===== Leverage discussions from platforms like RobotForum to avoid common pitfalls. Whether it's selecting the right planetary gearbox or figuring out the optimal motor for each joint, community insights can be invaluable. === Programming and Control === ==== ROS (Robot Operating System) ==== Start with ROS for an extensive suite of tools for programming and control, suitable for managing complex robotic functions. ROS serves as a valuable abstraction for comprehending the diverse components within a robotics system. However, its substantial nature may occasionally present challenges for modification for some users. The learning curve could be steep, and due to its emphasis on leveraging third-party packages, addressing issues can require additional effort and expertise. ==== Custom Software Solutions ==== Explore custom algorithms for adaptive control or reactive behaviors. Integrate advanced sensor feedback loops for real-time adjustments. == Experimenting with Your Humanoid Robot == === Testing and Iteration === ==== Virtual Testing Before Physical Implementation in Humanoid Robotics ==== NVIDIA's Isaac Sim and Isaac Gym, alongside other simulators, form a crucial foundation for designing and testing humanoid robots virtually. Insights and suggestions from experts working with these tools are captured below. ===== Isaac-Based Simulators and Frameworks ===== ====== Isaac Sim ====== IDE Experience: Provides a comprehensive, if complex, simulation environment. PhysX Engine: Utilizes the PhysX engine to handle both contact and joint constraints, though Isaac Sim currently does not fully expose closed-loop constraint capabilities. Joint Constraints: Supports maximal coordinate systems, which include joint constraints that are common in articulated robots. Virtual Sensors: Allows the simulation of perception with virtual cameras and LiDARs, providing policy training inputs rendered with NVIDIA RTX. ====== Isaac Gym ====== [[Reinforcement Learning]] Training: Enables parallel environments for fast policy training. PHC Approach: Integrates AMP for real-time pose control, making it easier to teach new skills. Gait Optimization Issues: While 17-DOF walking tasks work well, gait reward optimization needs refinement for more complex tasks. Closed-Loop Articulation: Belt-driven mechanisms provide a viable alternative for certain closed-loop designs. ====== Orbit Framework ====== Unified Training Framework: Integrates Isaac Sim and Isaac Gym for modular and consistent policy validation. OmniIsaacGymEnvs: Offers predefined tasks like walking and standing. ====== Omniverse Isaac Gym ====== Shift in Development: NVIDIA is consolidating Isaac Gym into Isaac Sim through Omniverse, providing the best of both worlds. Challenges: Demands powerful NVIDIA GPUs, potentially limiting some development workflows. ===== External Tools and Comparative Platforms ===== ====== Legged Gym ====== A repository showcasing the state-of-the-art in legged robot training. ====== MuJoCo (MJX)====== Offers a lightweight open-source alternative, supporting maximal coordinate constraints and easier to work with. The MuJoCo_MPC repository, created by Google DeepMind, is a toolset that combines Model Predictive Control (MPC) with the [[MuJoCo]] physics engine to create real-time behavior synthesis. With the advanced MJX extension, which uses GPU acceleration, it can simulate multiple environments in parallel. One approach is to try to replicate the techniques detailed in the AMP (Adversarial Motion Priors) paper to achieve agile humanoid behavior. For example, implementing a humanoid get-up sequence, which matches what was described in the AMP research. There’s been collaboration between different projects, like Stompy, to get humanoid simulations up and running. You could try converting Gymnasium to handle the URDF (Universal Robot Description Format) file format. Although converting to MJCF (MuJoCo's XML-based format) may present some challenges, we can still get it to work and refine the motor and actuator setup. Although [[MuJoCo]] can be slower in single-environment simulations, the MJX extension and its parallel processing potential make it a solid competitor. Compared to enviroments, like NVIDIA's Isaac Gym, MuJoCo might stand out for its extensibility and rapid development. One goal could be to try recreate the walking, running, and getting-up behaviors described in the AMP paper and use them as a foundation for training robust humanoid movements in simulation. ====== VSim ====== Claims to be 10x faster than other simulators. ====== ManiSkill/Sapien ====== Provides tactile simulation and visual-based policy training that is up to 100x faster than Isaac Sim. ===== Best Practices for Virtual Testing ===== - Incremental Complexity: Start simple and build up to more complex environments and tasks. - Cross-Simulator Validation: Validate robot models across simulators (e.g., Isaac and MuJoCo) to ensure robustness. - Incorporate Real-World Fidelity: Include sensor noise and imperfections for better policy generalization. - Optimize Resources: Use Azure's A100 GPUs for Isaac training. Capture real-world data to refine virtual training. By understanding the nuances and strengths of each simulator, developers can refine their humanoid robots effectively. Using Isaac Sim, Isaac Gym, and complementary tools, a robust simulation approach ensures smooth virtual-to-physical transferability while reducing development time and costs. More resources are available at [https://humanoids.wiki/w/Learning_algorithms Learning Algorithms] == Real-World Testing == Gradually transition to physical testing, beginning with simple tasks and moving to more complex interactions. === Data Collection and Analysis === ==== Camera Systems ==== Consider integrating advanced camera systems like those from e-con Systems or Arducam for visual feedback and navigation. Discuss camera choices, considering factors like latency, resolution, and integration ease with your main control system. == Advanced Customization and Community Engagement == === Open Source Projects === Contribute to an open-source project. For instance, platforms like GitHub host numerous projects where you can collaborate with others such as K-Scale https://github.com/kscalelabs. Check out [https://humanoids.wiki/w/Stompy Stompy!] === Modular Design === Engage in modular robot design to easily swap components or aesthetics. This approach allows for extensive customization and upgrades over time. == Safety and Continuous Learning == === Safety Protocols === Always implement robust safety measures when testing and demonstrating your robot. 8f43fe92e78ff148255f66e0395a2dfc29e5b380 4 17 1064 1036 2024-05-16T20:35:36Z Vrtnis 21 /*Cleaned up offtopic text*/ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. 10f836d98c345d22b9438c434524774f3df8d0d8 5 230 1065 984 2024-05-16T20:37:32Z Vrtnis 21 /* Removed offtopic talk text */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. 10f836d98c345d22b9438c434524774f3df8d0d8 6 248 1069 2024-05-17T07:33:58Z Ben 2 wikitext text/x-wiki Kscale db9 can bus convention 16e4691a83af1b3693cef81638b2e16b6b7f9379 6 249 1070 2024-05-17T07:34:26Z Ben 2 wikitext text/x-wiki Kscale phoenix can bus convention 3ffa0fc846f492ced9a6395794a3415bc9b4955d 0 2 1071 981 2024-05-17T07:35:06Z Ben 2 /* Wiring and Connectors */ wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions === Conventions === The images below show our pin convention for the CAN bus when using various connectors. <gallery> Kscale db9 can bus convention.jpg Kscale phoenix can bus convention.jpg </gallery> = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. = Artwork = Here's some art of Stompy! <gallery> Stompy 1.png Stompy 2.png Stompy 3.png Stompy 4.png </gallery> [[Category:Robots]] [[Category:Open Source]] [[Category:K-Scale]] c96e0188790ef7a2dade8c0ea1c8e1a83511c597 6 250 1076 2024-05-19T16:43:43Z Vrtnis 21 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 251 1084 2024-05-19T17:40:33Z Vrtnis 21 Created page with "=== Importing the Stompy Robot XML === To import the Stompy robot XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from..." wikitext text/x-wiki === Importing the Stompy Robot XML === To import the Stompy robot XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from the [https://github.com/google-deepmind/mujoco/releases/ GitHub repository] 2. '''Download Robot Files''': Obtain the Stompy robot MuJo files from [https://media.kscale.dev/stompy/latest_mjcf.tar.gz here]. 3. '''Drag and Drop''': Open the MuJoCo simulate binary. Simply drag and drop the `robot.xml` file from the downloaded robot files into the simulate window. e8948b51a6e0ff812f0ad44e4527e14fc8847070 1086 1084 2024-05-19T17:43:34Z Vrtnis 21 wikitext text/x-wiki === Importing the Stompy Robot XML === To import the Stompy robot XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from the [https://github.com/google-deepmind/mujoco/releases/ GitHub repository] 2. '''Download Robot Files''': Obtain the Stompy robot MuJo files from [https://media.kscale.dev/stompy/latest_mjcf.tar.gz here]. 3. '''Drag and Drop''': Open the MuJoCo simulate binary. Simply drag and drop the `robot.xml` file from the downloaded robot files into the simulate window. [[File:Mujucoreleases1.png|300px|thumb|none|MuJoCo releases]] 5abe433883e7787ae6f2f730f99db780df34eeb6 1088 1086 2024-05-19T17:49:37Z Vrtnis 21 wikitext text/x-wiki === Importing the Stompy Robot XML === To import the Stompy robot XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from the [https://github.com/google-deepmind/mujoco/releases/ GitHub repository] [[File:Mujucoreleases1.png|600px|thumb|none|MuJoCo releases]] 2. '''Download Robot Files''': Obtain the Stompy robot MuJo files from [https://media.kscale.dev/stompy/latest_mjcf.tar.gz here]. [[File:Storekscalemjcf.png|600px|thumb|none| K-Scale Downloads]] 3. '''Drag and Drop''': Open the MuJoCo simulate binary. Simply drag and drop the `robot.xml` file from the downloaded robot files into the simulate window. 18a4ed35c42631af5b34933505f494d47f7d478f 1091 1088 2024-05-19T18:00:45Z Vrtnis 21 wikitext text/x-wiki === Importing the Stompy Robot XML === To import the Stompy robot XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from the [https://github.com/google-deepmind/mujoco/releases/ GitHub repository] [[File:Mujucoreleases1.png|600px|thumb|none|MuJoCo releases]] 2. '''Download Robot Files''': Obtain the Stompy robot MuJoCo files from [https://media.kscale.dev/stompy/latest_mjcf.tar.gz here]. [[File:Storekscalemjcf.png|600px|thumb|none| K-Scale Downloads]] 3. '''Drag and Drop''': Open the MuJoCo simulate binary. Simply drag and drop the `robot.xml` file from the downloaded robot files into the simulate window. [[File:Run simulate.png|600px|frame|none|Run MuJoCo Simulate]] [[File:Robotxml.png|600px|thumb|none|Downloaded Files]] d37c9e11f91e877f7154b61c09caebb3ce49aa8d 1093 1091 2024-05-19T18:05:39Z Vrtnis 21 /*Add MuJoCo screenshots*/ wikitext text/x-wiki === Importing the Stompy Robot XML === To import the Stompy robot XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from the [https://github.com/google-deepmind/mujoco/releases/ GitHub repository] [[File:Mujucoreleases1.png|600px|thumb|none|MuJoCo releases]] 2. '''Download Robot Files''': Obtain the Stompy robot MuJoCo files from [https://media.kscale.dev/stompy/latest_mjcf.tar.gz here]. [[File:Storekscalemjcf.png|600px|thumb|none| K-Scale Downloads]] 3. '''Drag and Drop''': Open the MuJoCo simulate binary. Simply drag and drop the `robot.xml` file from the downloaded robot files into the simulate window. [[File:Run simulate.png|600px|frame|none|Run MuJoCo Simulate]] [[File:Robotxml.png|600px|thumb|none|Downloaded Files]] [[File:Stompy drag drop mujoco.png|800px|thumb|none|Drop into MuJoCo Simulate]] fb779379b0f74d1c77f2f2f259ff42159aaa7e11 1094 1093 2024-05-19T18:06:52Z Vrtnis 21 /*Added Guides Tag*/ wikitext text/x-wiki === Importing the Stompy Robot XML === To import the Stompy robot XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from the [https://github.com/google-deepmind/mujoco/releases/ GitHub repository] [[File:Mujucoreleases1.png|600px|thumb|none|MuJoCo releases]] 2. '''Download Robot Files''': Obtain the Stompy robot MuJoCo files from [https://media.kscale.dev/stompy/latest_mjcf.tar.gz here]. [[File:Storekscalemjcf.png|600px|thumb|none| K-Scale Downloads]] 3. '''Drag and Drop''': Open the MuJoCo simulate binary. Simply drag and drop the `robot.xml` file from the downloaded robot files into the simulate window. [[File:Run simulate.png|600px|frame|none|Run MuJoCo Simulate]] [[File:Robotxml.png|600px|thumb|none|Downloaded Files]] [[File:Stompy drag drop mujoco.png|800px|thumb|none|Drop into MuJoCo Simulate]] [[Category: Guides]] c5f2e2626cb124bf5bb84ae4f4a21581f96b50a9 1095 1094 2024-05-19T18:14:55Z Vrtnis 21 wikitext text/x-wiki === Importing the Stompy Robot XML === To import Stompy's XML into MuJoCo, follow these steps: 1. '''Download the Binary''' : First, download the MuJoCo binary from the [https://github.com/google-deepmind/mujoco/releases/ GitHub repository] [[File:Mujucoreleases1.png|600px|thumb|none|MuJoCo releases]] 2. '''Download Robot Files''': Obtain the Stompy robot MuJoCo files from [https://media.kscale.dev/stompy/latest_mjcf.tar.gz here]. [[File:Storekscalemjcf.png|600px|thumb|none| K-Scale Downloads]] 3. '''Drag and Drop''': Open the MuJoCo simulate binary. Simply drag and drop the `robot.xml` file from the downloaded robot files into the simulate window. [[File:Run simulate.png|600px|frame|none|Run MuJoCo Simulate]] [[File:Robotxml.png|600px|thumb|none|Downloaded Files]] [[File:Stompy drag drop mujoco.png|800px|thumb|none|Drop into MuJoCo Simulate]] [[Category: Guides]] 5df30072c630cdb52b23658bccf0ee20c03966be 6 252 1085 2024-05-19T17:42:44Z Vrtnis 21 wikitext text/x-wiki Screenshot of MuJoCo releases 5c8406c0bc12550eb294e17ee0bc38ed53955515 6 253 1087 2024-05-19T17:48:05Z Vrtnis 21 wikitext text/x-wiki KScale MJCF Download 37caa6cccc01c52c6e4026f53ce363bc8b50facd 6 254 1089 2024-05-19T17:51:37Z Vrtnis 21 wikitext text/x-wiki Robot XML Download 75f74e78599c7066712bfedf574b24334e8f8b81 6 255 1090 2024-05-19T17:59:48Z Vrtnis 21 wikitext text/x-wiki Run ./simulate 44535584d6cb4bfd412c2741e20745363df98de6 6 256 1092 2024-05-19T18:04:07Z Vrtnis 21 wikitext text/x-wiki Drag and Drop Into MuJoCo Simulate c0166c795dc626e6efe1ae8ff77fa78a4e78a8d9 0 257 1097 2024-05-19T18:58:40Z Vrtnis 21 /*Setup guide for MuJoCo WASM*/ wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 1098 1097 2024-05-19T19:01:32Z Vrtnis 21 /*Add install process*/ wikitext text/x-wiki == Building == === 1. Install emscripten === First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === 2. Build the mujoco_wasm Binary === Next, you'll build the MuJoCo WebAssembly binary. ==== On Linux ==== <code> mkdir build cd build emcmake cmake .. make </code> 129397f418eac4b0516b594b002de9e882441f17 1099 1098 2024-05-19T19:15:55Z Vrtnis 21 /*emscripten instructions*/ wikitext text/x-wiki == Building == === 1. Install emscripten === First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === 1. Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === 2. Enter that directory === <code> cd emsdk </code> === 3. Download and install the latest SDK tools === <code> ./emsdk install latest </code> === 4. Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === 5. Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === 6. Now just try it! === <code> emcc </code> === 2. Build the mujoco_wasm Binary === Next, you'll build the MuJoCo WebAssembly binary. ==== On Linux ==== <code> mkdir build cd build emcmake cmake .. make </code> 59e77c2b3b8841f7edd7ba28aa0b833ad52e20b9 1100 1099 2024-05-19T19:31:16Z Vrtnis 21 /*Update build instructions*/ wikitext text/x-wiki == Building == === 1. Install emscripten === First, you need to install emscripten, which is a compiler toolchain for WebAssembly. ==== 1.1. Get the emsdk repo ==== <code> git clone https://github.com/emscripten-core/emsdk.git </code> ==== 1.2. Enter that directory ==== <code> cd emsdk </code> ==== 1.3. Download and install the latest SDK tools ==== <code> ./emsdk install latest </code> ==== 1.4. Make the "latest" SDK "active" ==== <code> ./emsdk activate latest </code> ==== 1.5. Activate PATH and other environment variables ==== <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> ==== 1.6. Now just try it! ==== <code> emcc </code> '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> === 2. Build the mujoco_wasm Binary === Next, you'll build the MuJoCo WebAssembly binary. ==== On Linux ==== <code> mkdir build cd build emcmake cmake .. make </code> 15611c14f8d5a5327dddbab31d26e38663adb57c 1101 1100 2024-05-19T19:40:53Z Vrtnis 21 wikitext text/x-wiki == Building == === 1. Install emscripten === First, you need to install emscripten, which is a compiler toolchain for WebAssembly. ==== Get the emsdk repo ==== <code> git clone https://github.com/emscripten-core/emsdk.git </code> ==== Enter that directory ==== <code> cd emsdk </code> ==== Download and install the latest SDK tools ==== <code> ./emsdk install latest </code> ==== Make the "latest" SDK "active" ==== <code> ./emsdk activate latest </code> ==== Activate PATH and other environment variables ==== <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> ==== Now just try it! ==== <code> emcc </code> '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> === 2. Build the mujoco_wasm Binary === Next, you'll build the MuJoCo WebAssembly binary. ==== On Linux ==== <code> mkdir build cd build emcmake cmake .. make </code> === 2. Build the mujoco_wasm Binary === Next, you'll build the MuJoCo WebAssembly binary. ==== On Linux ==== <code> mkdir build cd build emcmake cmake .. make </code> 2570e901554e22950822332c612c9e83c849eed3 1102 1101 2024-05-19T19:46:37Z Vrtnis 21 wikitext text/x-wiki == 1. Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === 1.1. Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === 1.2. Enter that directory === <code> cd emsdk </code> === 1.3. Download and install the latest SDK tools === <code> ./emsdk install latest </code> === 1.4. Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === 1.5. Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === 1.6. Now just try it! === <code> emcc </code> '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == 2. Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. === 2.1. On Linux === <code> mkdir build cd build emcmake cmake .. make </code> 8579635af0b02dbdd905ee5e8bce79085502c074 1103 1102 2024-05-19T19:47:49Z Vrtnis 21 /*Reorg*/ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. === On Linux === <code> mkdir build cd build emcmake cmake .. make </code> 3d3dacfcc88606f75022c02bfc23ea4be1af8737 1104 1103 2024-05-19T19:49:01Z Vrtnis 21 wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <code> mkdir build cd build emcmake cmake .. make </code> '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> a5fbe0bb4b5784d62cd08db011deb4d06d3551c2 1105 1104 2024-05-19T20:00:51Z Vrtnis 21 /* Build the mujoco_wasm Binary */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> 4b7e66d0975441a40ff55c9892af7145d0439181 1107 1105 2024-05-19T20:20:03Z Vrtnis 21 /* Add screenshot */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Emcmake.png|800px|thumb|none|emcmake cmake ..]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> e5cd7387c517b787275df96a8ae1d75d9404c5ec 6 258 1106 2024-05-19T20:19:24Z Vrtnis 21 wikitext text/x-wiki emcmake 86248b08caa9aaab9ac370c07d55f80e27e5088e 6 259 1108 2024-05-19T20:23:51Z Vrtnis 21 wikitext text/x-wiki wasm build bff69fa2a678fd537b7ca33a54c1263fffb45ed7 0 257 1109 1107 2024-05-19T20:24:49Z Vrtnis 21 /* Build the mujoco_wasm Binary */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> bc5c1125d587c1a6bef1ee3032041aad6dadc0eb 1111 1109 2024-05-19T20:27:20Z Vrtnis 21 /* Build the mujoco_wasm Binary */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> fc1a06fcc2c170c294189b1396faf2ff437e7ac1 1112 1111 2024-05-19T20:33:03Z Vrtnis 21 wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Running in Browser == Now just do in the root of your mujoco folder <code> python -m http.server 8000 </code> Then navigate to: <code> http://localhost:8000/index.html </code> 03949e76f1b818243e4028cbb856acb25f58928c 1113 1112 2024-05-19T20:33:24Z Vrtnis 21 /* Running in Browser */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Running in Browser == Now just do in the root of your mujoco folder <code> python -m http.server 8000 </code> Then navigate to: <code> http://localhost:8000/index.html </code> 1fdc8f30da406f339003663d9da7b4bf1d536131 1115 1113 2024-05-19T20:42:03Z Vrtnis 21 /* Running in Browser */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Running in Browser == Now just do in the root of your mujoco folder <code> python -m http.server 8000 </code> Then navigate to: <code> http://localhost:8000/index.html </code> [[File:Wasm screenshot13-40-40.png|800px|thumb|none|MuJoCo running in browser]] 7caaa8311c256a957122e067e04d4612687a3a44 1116 1115 2024-05-19T21:24:47Z Vrtnis 21 /* Running in Browser */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Running in Browser == Run this in your mujoco folder to start a server. <code> python -m http.server 8000 </code> Then navigate to: <code> http://localhost:8000/index.html </code> [[File:Wasm screenshot13-40-40.png|800px|thumb|none|MuJoCo running in browser]] 84106a7907262faa62db2a6ddf56c528e51ce230 6 260 1110 2024-05-19T20:26:40Z Vrtnis 21 wikitext text/x-wiki make 5821eb27d7b71c9078000da31a5a654c97e401b9 6 261 1114 2024-05-19T20:41:22Z Vrtnis 21 wikitext text/x-wiki mujoco wasm screenshot a53efe8f5741b3111faa2da0556dad3a0022dc66 2 262 1117 2024-05-19T23:39:04Z Is2ac 30 Created page with " {{infobox person | name = Isaac Light | organization = [[K-Scale Labs]] | title = Employee }} [[Category: K-Scale Employees]]" wikitext text/x-wiki {{infobox person | name = Isaac Light | organization = [[K-Scale Labs]] | title = Employee }} [[Category: K-Scale Employees]] dd118feb47faa916a98a790323063aaca8b09b2d 0 263 1118 2024-05-20T01:29:17Z Ben 2 Created page with "This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] **..." wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode 62db86cfd15f4ab215f53ee2dc42a3d569861926 1119 1118 2024-05-20T01:41:02Z Ben 2 wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code 75ead204c219c5c1eaddc5ebc369352857364813 1120 1119 2024-05-20T01:47:14Z Ben 2 wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code * Check out the [https://github.com/kscalelabs/mlfab/tree/master/examples mlfab examples directory] for some ideas * It is a good idea to try to write the full training loop yourself to figure out what's going on f31225302799a017de2cd1d52c40d956c2880e9f 1126 1120 2024-05-20T03:11:12Z Dennisc 27 wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code * Check out the [https://github.com/kscalelabs/mlfab/tree/master/examples mlfab examples directory] for some ideas * It is a good idea to try to write the full training loop yourself to figure out what's going on * Run `nvidia-smi` to see the GPUs and their statuses/any active processes 722041ea6c7159b5982c148a4deb7f54f135c579 1127 1126 2024-05-20T03:11:30Z Dennisc 27 wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code * Check out the [https://github.com/kscalelabs/mlfab/tree/master/examples mlfab examples directory] for some ideas * It is a good idea to try to write the full training loop yourself to figure out what's going on * Run <code>nvidia-smi</code> to see the GPUs and their statuses/any active processes 48682cf497d4672309f1c78908fc1aa8b51b6c91 1128 1127 2024-05-20T03:19:45Z Dennisc 27 skeleton of uv wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 ==== uv ==== To get started with <code>uv</code>, pick a directory you want your virtual environment to live in. (<code>$HOME</code> is not recommended.) Once you have <code>cd</code>ed there, run <syntaxhighlight lang="bash"> uv venv </syntaxhighlight> To activate your virtual environment, run <syntaxhighlight lang="bash"> source .venv/bin/activate </syntaxhighlight> *while in the directory you created your <code>.venv</code> in*. === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code * Check out the [https://github.com/kscalelabs/mlfab/tree/master/examples mlfab examples directory] for some ideas * It is a good idea to try to write the full training loop yourself to figure out what's going on * Run <code>nvidia-smi</code> to see the GPUs and their statuses/any active processes fda062259e6fccf9c4d3ef5563c4d1018ae1cdad 1129 1128 2024-05-20T03:20:08Z Dennisc 27 fix bold wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 ==== uv ==== To get started with <code>uv</code>, pick a directory you want your virtual environment to live in. (<code>$HOME</code> is not recommended.) Once you have <code>cd</code>ed there, run <syntaxhighlight lang="bash"> uv venv </syntaxhighlight> To activate your virtual environment, run <syntaxhighlight lang="bash"> source .venv/bin/activate </syntaxhighlight> '''while in the directory you created your <code>.venv</code> in'''. === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code * Check out the [https://github.com/kscalelabs/mlfab/tree/master/examples mlfab examples directory] for some ideas * It is a good idea to try to write the full training loop yourself to figure out what's going on * Run <code>nvidia-smi</code> to see the GPUs and their statuses/any active processes 70ded971b6eeee4d1b667011b3ec82e7df254c1a 1130 1129 2024-05-20T03:23:58Z Dennisc 27 How to install uv on cluster wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 ==== uv ==== To install <code>uv</code> on the K-Scale clusters, run <syntaxhighlight lang="bash"> curl -LsSf https://astral.sh/uv/install.sh | sh </syntaxhighlight> To get started with <code>uv</code>, pick a directory you want your virtual environment to live in. (<code>$HOME</code> is not recommended.) Once you have <code>cd</code>ed there, run <syntaxhighlight lang="bash"> uv venv </syntaxhighlight> To activate your virtual environment, run <syntaxhighlight lang="bash"> source .venv/bin/activate </syntaxhighlight> '''while in the directory you created your <code>.venv</code> in'''. === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code * Check out the [https://github.com/kscalelabs/mlfab/tree/master/examples mlfab examples directory] for some ideas * It is a good idea to try to write the full training loop yourself to figure out what's going on * Run <code>nvidia-smi</code> to see the GPUs and their statuses/any active processes c1414a6754df6d02dda2b53e9cb9f3b28a967cda 1131 1130 2024-05-20T03:33:05Z Dennisc 27 Explain uv venv --python 3.11 flag and why it is important/why one might want to use it. wikitext text/x-wiki This is [[User:Ben]]'s guide to getting started with machine learning. === Dependencies === Here's some useful dependencies that I use: * [https://astral.sh/blog/uv uv] ** This is similar to Pip but written in Rust and is way faster ** It has nice management of virtual environments ** Can use Conda instead but it is much slower * [https://github.com/features/copilot Github Copilot] * [https://github.com/kscalelabs/mlfab mlfab] ** This is a Python package I made to help make it easy to quickly try out machine learning ideas in PyTorch * Coding tools ** [https://mypy-lang.org/ mypy] static analysis ** [https://github.com/psf/black black] code formatter ** [https://docs.astral.sh/ruff/ ruff] alternative to flake8 ==== uv ==== To install <code>uv</code> on the K-Scale clusters, run <syntaxhighlight lang="bash"> curl -LsSf https://astral.sh/uv/install.sh | sh </syntaxhighlight> To get started with <code>uv</code>, pick a directory you want your virtual environment to live in. (<code>$HOME</code> is not recommended.) Once you have <code>cd</code>ed there, run <syntaxhighlight lang="bash"> uv venv </syntaxhighlight> '''If you are on the clusters''', you instead may want to run <syntaxhighlight lang="bash"> uv venv --python 3.11 </syntaxhighlight> to ensure that the virtual environment uses Python 3.11. This is because by default, uv uses the system's version of Python (whatever the result of <code>which python</code> yields), and the clusters are running Python 3.10.12. (Python 3.11 will be useful because various projects, including the starter project, will require Python 3.11.) To activate your virtual environment, run <syntaxhighlight lang="bash"> source .venv/bin/activate </syntaxhighlight> '''while in the directory you created your <code>.venv</code> in'''. === Installing Starter Project === * Go to [https://github.com/kscalelabs/getting-started this project] and install it ==== Opening the project in VSCode ==== * Create a VSCode config file that looks something like this: <syntaxhighlight lang="json"> { "folders": [ { "name": "Getting Started", "path": "/home/ubuntu/Github/getting_started" }, { "name": "Workspaces", "path": "/home/ubuntu/.code-workspaces" } ], "settings": { "cmake.configureSettings": { "CMAKE_CUDA_COMPILER": "/usr/bin/nvcc", "CMAKE_PREFIX_PATH": [ "/home/ubuntu/.virtualenvs/getting-started/lib/python3.11/site-packages/torch/share/cmake" ], "PYTHON_EXECUTABLE": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "TORCH_CUDA_ARCH_LIST": "'8.0'" }, "python.defaultInterpreterPath": "/home/ubuntu/.virtualenvs/getting-started/bin/python", "ruff.path": [ "/home/ubuntu/.virtualenvs/getting-started/bin/ruff" ] } } </syntaxhighlight> * Install the [https://code.visualstudio.com/docs/remote/ssh VSCode SSH extension] * SSH into the cluster (see [[K-Scale Cluster]] for instructions) * Open the workspace that you created in VSCode === Useful Brain Dump Stuff === * Use <code>breakpoint()</code> to debug code * Check out the [https://github.com/kscalelabs/mlfab/tree/master/examples mlfab examples directory] for some ideas * It is a good idea to try to write the full training loop yourself to figure out what's going on * Run <code>nvidia-smi</code> to see the GPUs and their statuses/any active processes 7aebade2c72b449559a1769fc75d0cd5689410ee 0 1 1121 1079 2024-05-20T02:41:17Z Modeless 7 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Haier]] | |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} bcaf875e1a5c513b2b09f73db22b75ede78deca0 0 264 1122 2024-05-20T02:46:39Z Modeless 7 Created page with "Haier is a Chinese manufacturer of appliances and consumer electronics. There's [a video of a Haier-branded humanoid here](https://www.youtube.com/watch?v=7Ve9UTUJZG8), althou..." wikitext text/x-wiki Haier is a Chinese manufacturer of appliances and consumer electronics. There's [a video of a Haier-branded humanoid here](https://www.youtube.com/watch?v=7Ve9UTUJZG8), although there are no further details. 1b3a9eca71600d4d908d475d026d8fb815f5b371 1123 1122 2024-05-20T02:47:15Z Modeless 7 wikitext text/x-wiki Haier is a Chinese manufacturer of appliances and consumer electronics. There's [[a video of a Haier-branded humanoid here|https://www.youtube.com/watch?v=7Ve9UTUJZG8]], although there are no further details. 7908355773ae8ec82b2d850bbd381bbaea3efe4f 1124 1123 2024-05-20T02:47:27Z Modeless 7 wikitext text/x-wiki Haier is a Chinese manufacturer of appliances and consumer electronics. There's [[https://www.youtube.com/watch?v=7Ve9UTUJZG8|a video of a Haier-branded humanoid here]], although there are no further details. f56cbdabdf3c8f734fb876a3fa6354d07b615e74 1125 1124 2024-05-20T02:47:38Z Modeless 7 wikitext text/x-wiki Haier is a Chinese manufacturer of appliances and consumer electronics. There's [https://www.youtube.com/watch?v=7Ve9UTUJZG8|a video of a Haier-branded humanoid here], although there are no further details. 13101097ee30704d2ec578cba5809df75f97d22d 0 265 1132 2024-05-20T19:46:21Z Ben 2 Created page with "This document contains a walk-through of how to set up MediaWiki on AWS. In total, this process should take about 30 minutes. === Getting Started === # Use [https://aws.amaz..." wikitext text/x-wiki This document contains a walk-through of how to set up MediaWiki on AWS. In total, this process should take about 30 minutes. === Getting Started === # Use [https://aws.amazon.com/marketplace/pp/prodview-3tokjpxwvddp2 this service] to install MediaWiki to an EC2 instance # After installing, SSH into the instance. The MediaWiki root files live in <code>/var/www/html</code>, and Apache configuration files live in <code>/etc/apache2</code> === Set up A Record === In your website DNS, create an A record which points from your desired URL to the IP address of the newly-created EC2 instance. For example: * '''Host name''': <code>@</code> (for the main domains) or <code>wiki</code> (for subdomains) * '''Type''': <code>A</code> * '''Data''': <code>127.0.0.1</code> (use the IP address of the EC2 instance) === Installing OpenSSL === Roughly follow [https://www.digitalocean.com/community/tutorials/how-to-secure-apache-with-let-s-encrypt-on-ubuntu-20-04 these instructions] to install OpenSSL using LetsEncrypt # Install certbot <syntaxhighlight lang="bash"> sudo apt install certbot python3-certbot-apache </syntaxhighlight> # Run certbot (you can just select "No redirect" <syntaxhighlight lang="bash"> sudo certbot --apache </syntaxhighlight> # Verify certbot <syntaxhighlight lang="bash"> sudo systemctl status certbot.timer </syntaxhighlight> === Update Page URLs === To get MediaWiki to display pages using the <code>/w/Some_Page</code> URL, modify <code>LocalSettings.php</code>. Replace the line where it says <code>$wgScriptPage = "";</code> with the following: <syntaxhighlight lang="php"> $wgScriptPath = ""; $wgArticlePath = "/w/$1"; $wgUsePathInfo = true; </syntaxhighlight> Next, in <code>/etc/apache2/apache2.conf</code>, add the following section: <syntaxhighlight lang="text"> <Directory /var/www/html> AllowOverride all </Directory> </syntaxhighlight> Create a new <code>.htaccess</code> in <code>/var/www/html</code> with the following contents: <syntaxhighlight lang="text"> RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule ^w/(.*)$ index.php/$1 [PT,L,QSA] </syntaxhighlight> Finally, reload Apache2: <syntaxhighlight lang="bash"> sudo service apache2 reload </syntaxhighlight> === Extras === Here are some notes about some extra features you can add on top of MediaWiki. ==== Storing images in S3 with Cloudfront ==== You can use the [https://www.mediawiki.org/wiki/Extension:AWS AWS MediaWiki Extension] to allow users to upload images and files, store them in S3, and serve them through CloudFront. This process is somewhat involved. I may write notes about how to do this well in the future. d03b04998fef32b797f4648ec7854c04b05dea5b 0 266 1133 2024-05-20T19:54:28Z Vrtnis 21 Created page with "[[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover)" wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) fe184d1f26022168d3e1225e568c8289f6ab8b5e 1134 1133 2024-05-20T19:55:12Z Vrtnis 21 /*mars rover */ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. 44f6a96d1f447ff4f9ad32e5a9fdd77581a2ff5f 1135 1134 2024-05-20T19:56:54Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. === risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. c91c0f3351de90272bb45c5889ddd57760da98e8 1136 1135 2024-05-20T19:57:02Z Vrtnis 21 /* risk-aware planning */ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. cdcba16b8c3150efaabcf4c0db35fc3273d155c1 1137 1136 2024-05-20T19:57:07Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. 4134a20b0eccf641b326fa0f2ab65f03a2341c7e 1139 1137 2024-05-20T20:00:51Z Vrtnis 21 /*add slide*/ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. fcc5677dfc8747e7c8d1018291086411a0a1d7c2 1140 1139 2024-05-20T20:02:24Z Vrtnis 21 /*sensor integration*/ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. === Advanced sensor integration === * high-resolution cameras and spectrometers. * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). 18d4fa33c088111e0adefde2dbfb7ebc83ae32b5 1141 1140 2024-05-20T20:02:33Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. === Advanced sensor integration === * high-resolution cameras and spectrometers. * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). ddbe914e20dfde5db892802ae4254910fdcc22f0 1142 1141 2024-05-20T20:14:55Z Vrtnis 21 /*Energy management and path planning*/ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. === Advanced sensor integration === * high-resolution cameras and spectrometers. * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning,. * integrate real-time obstacle detection 6416903142cc40bb05633038143b837e8330b3fb 1143 1142 2024-05-20T20:15:05Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. === Advanced sensor integration === * high-resolution cameras and spectrometers. * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning,. * integrate real-time obstacle detection 6df279475a4a24a8d342ac425f97b1dc4f4eafb5 1145 1143 2024-05-20T20:17:39Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers. * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning,. * integrate real-time obstacle detection db31e3e81a07dad370ca441349575bb9f2fff97d 1146 1145 2024-05-20T20:28:58Z Vrtnis 21 /*project management lessons*/ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner to navigate safely in unstructured environments. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * engineer the autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers. * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning,. * integrate real-time obstacle detection === Project managementlessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts 35b7b6bc836fbed3ab7060745e0d4aa10bc587e0 1147 1146 2024-05-20T20:29:47Z Vrtnis 21 /* Mars rover autonomy */ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers. * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning,. * integrate real-time obstacle detection === Project managementlessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts 980473f281ea1e03d1990fd9361cbf49970657fe 1148 1147 2024-05-20T20:30:06Z Vrtnis 21 /* Advanced sensor integration */ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning,. * integrate real-time obstacle detection === Project managementlessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts 261c544790f9d7c655e8d5b1634f58f4b5d3182e 1149 1148 2024-05-20T22:16:47Z Vrtnis 21 /* Navigation and mapping */ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning * integrate real-time obstacle detection === Project managementlessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts 14cc13a42efeb0e96cc04f527c3953f911befd4e 1150 1149 2024-05-20T22:19:40Z Vrtnis 21 /* Project managementlessons learned */ wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning * integrate real-time obstacle detection === Project management lessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts a816bd3bc6d369127f4dde4a31c8502e717fd713 1151 1150 2024-05-20T22:25:04Z Vrtnis 21 Vrtnis moved page [[JPL Robotics Lessons Learnt]] to [[JPL Robotics Meeting Notes]]: Correct title wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning * integrate real-time obstacle detection === Project management lessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts a816bd3bc6d369127f4dde4a31c8502e717fd713 1153 1151 2024-05-20T22:48:32Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Humanoids === * for space exploration, utilizing human-like structure * tasks requiring precise human-like movements. * deploy humanoids for maintenance, repair enhancing astronaut safety. * use in sample gathering/ environmental monitoring. * simulate operations in space-like environments. === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning * integrate real-time obstacle detection === Project management lessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts 5b904356b71b96e1297afb2bbcbd6d4a2f94b89a 1154 1153 2024-05-20T22:48:49Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance and safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain, improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Humanoids === * for space exploration, utilizing human-like structure * tasks requiring precise human-like movements. * deploy humanoids for maintenance, repair enhancing astronaut safety. * use in sample gathering/ environmental monitoring. * simulate operations in space-like environments. === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning * integrate real-time obstacle detection === Project management lessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts cfb48c0d66d26caeea9b7e850e829761c1cc1d2d 1155 1154 2024-05-20T22:50:33Z Vrtnis 21 wikitext text/x-wiki [[User:vrtnis|User:vrtnis]]' notes from meeting with JPL robotics head of robotics team (EELS and Rover) === Mars rover autonomy === * combine approximate kinematic settling with a dual-cost path planner. * develop multi-agent capabilities to significantly enhance performance - safety of autonomous operations on Mars. * autonomy system to efficiently handle Mars' unpredictable terrain - improving mission success rates. * incorporate mechanisms for self-diagnosis and repair to ensure long-term functionality with minimal Earth-based support. [[File:Screenshot12-58-42.png|800px|thumb|none|Summary]] === Risk-aware planning === * utilize Boole's inequality for risk allocation. * enhancing decision-making under uncertainty. * employ predictive analytics to anticipate and mitigate potential risks. * allowing for dynamic re-planning and risk management. [[File:Screenshot13-16-39.png|400px|thumb|none]] === Advanced sensor integration === * high-resolution cameras and spectrometers * integrate sensors for comprehensive environmental data. * using ground-penetrating radar (GPR). === Humanoids === * for space exploration, utilizing human-like structure * tasks requiring precise human-like movements. * deploy humanoids for maintenance, repair enhancing astronaut safety. * use in sample gathering/ environmental monitoring. * simulate operations in space-like environments. === Energy === * optimize energy consumption. * utilize a combination of solar panels. * gather power from the environment, such as thermal gradients and mechanical movements. === Navigation and mapping === * utilize LIDAR * optimal path planning * integrate real-time obstacle detection === Project management lessons learned === * manage complex robotics projects like eels and Mars rovers by coordinating multiple teams * advantage of having all team members colocated in the same building at the project's inception. * more integrated workflows through initial colocation * ensure alignment on project goals and timelines. * early stages of complex projects benefit greatly from in-person collaboration. establish a strong foundation through colocation to enhance subsequent remote coordination efforts e38a80c98642676587fae063a2f22deb53ac33ff 6 267 1138 2024-05-20T19:59:49Z Vrtnis 21 wikitext text/x-wiki /*summary slide*/ 993a32dde4daa7fad481a7b8654815a5ccf40796 6 268 1144 2024-05-20T20:17:18Z Vrtnis 21 wikitext text/x-wiki path planning slide c8c623e5024ac420ca9311b91cda1d348e931b9e 0 269 1152 2024-05-20T22:25:04Z Vrtnis 21 Vrtnis moved page [[JPL Robotics Lessons Learnt]] to [[JPL Robotics Meeting Notes]]: Correct title wikitext text/x-wiki #REDIRECT [[JPL Robotics Meeting Notes]] cd85744a4ee3eb5a360188d3ce1f77c9abae7e06 0 270 1156 2024-05-21T03:10:17Z Ben 2 Created page with "Allen's reinforcement learning notes === Links === * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture]" wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] 1723d954a8b80f71a0963544656871c99747c3d7 1157 1156 2024-05-21T03:10:32Z Ben 2 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] d248ef20442690b683cc0f041482e78b8190032b 0 34 1158 1056 2024-05-21T03:11:00Z Ben 2 wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == RL in Robotics == === Practical Applications of RL === ==== Task Automation ==== * Robots can be trained to perform repetitive or dangerous tasks autonomously, such as assembly line work, welding, or hazardous material handling. * RL enables robots to adapt to new tasks without extensive reprogramming, making them versatile for various industrial applications. ==== Navigation and Manipulation ==== * RL is used to train robots for navigating complex environments and manipulating objects with precision, which is crucial for tasks like warehouse logistics, domestic chores, and medical surgeries. === Simulation Environments === ==== Isaac Sim ==== * Isaac Sim provides a highly realistic and interactive environment where robots can be trained safely and efficiently. * The simulated environment includes physics, sensors, and other elements that mimic real-world conditions, enabling the transfer of learned behaviors to physical robots. ==== MuJoCo ==== * MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research and development in robotics, machine learning, and biomechanics. * It offers fast and accurate simulations, which are essential for training RL agents in tasks involving complex dynamics and contact-rich interactions. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] [[Category: Software]] [[Category: Reinforcement Learning]] e4c57d8dc0b824b4ad78aa7b0939188a09f57d8c 0 270 1159 1157 2024-05-21T03:11:49Z Ben 2 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] 5abb12a2472d8018339336e360522562fc482859 1161 1159 2024-05-21T05:43:47Z Allen12 15 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy pi that maps states or observations to actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t <\math> to maximize <math> \sum_t r_t <\math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a bayes net, we have * s_1 -> o_1 - (pi_theta) -> a_1 (policy) * s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. When choosing hyperparameters we need to be careful to make sure that we go for completing long term goals instead of always looking for immediate reward. == Markov Chain & Decision Process== Markov Chain: <math> M = {S, T} <\math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} <\math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k). r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here \alpha represents the learning rate, which is how much new information is weighted relative to old information, while \gamma represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. -- feb3c2c25a2c337385632f5c58f13996cfa33a84 1162 1161 2024-05-21T05:44:08Z Allen12 15 /* Markov Chain & Decision Process */ wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy pi that maps states or observations to actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t <\math> to maximize <math> \sum_t r_t <\math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a bayes net, we have * s_1 -> o_1 - (pi_theta) -> a_1 (policy) * s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. When choosing hyperparameters we need to be careful to make sure that we go for completing long term goals instead of always looking for immediate reward. === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} <\math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} <\math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k). r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here \alpha represents the learning rate, which is how much new information is weighted relative to old information, while \gamma represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. -- ece368361a6f395de83df2b2a0e062adf95ac799 1163 1162 2024-05-21T05:56:23Z Allen12 15 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy pi that maps states or observations to actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t <\math> to maximize <math> \sum_t r_t <\math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a bayes net, we have * s_1 -> o_1 - (pi_theta) -> a_1 (policy) * s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. When choosing hyperparameters we need to be careful to make sure that we go for completing long term goals instead of always looking for immediate reward. === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} <\math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} <\math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k). r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here \alpha represents the learning rate, which is how much new information is weighted relative to old information, while \gamma represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. b42a20c6f2a6141bf1b7a0eaac7649937dc504d7 1164 1163 2024-05-21T15:47:13Z 108.211.178.220 0 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy pi that maps states or observations to actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a network of conditional probability, we have * <math> s_1 -> o_1 - (pi_theta) -> a_1 </math> (policy) * <math> s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 </math> (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. We often use and tune hyperparameters for reward functions to make model training faster === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} </math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} </math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k). r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat Policy Gradients - Directly differentiate objective with respect to the optimal theta and then perform gradient descent Value-based: Estimate value function or q-function of optimal policy (policy is often represented implicitly) Actor-Critic: Estimate value function or q-function of current policy, and find a better policy gradient Model-based: Estimate some transition model, and then use it to improve a policy === REINFORCE === - === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here <math>\alpha</math> represents the learning rate, which is how much new information is weighted relative to old information, while <math>\gamma</math> represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. <math> Q_i+1(s,a)=(1−\alpha)Q_i(s,a)+\alpha(r(s, a)+\gammaV_i(s')) </math> Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. 1050fd88814976490830b1b30bb20856a8de0a22 1166 1164 2024-05-21T18:15:02Z Ben 2 /* Markov Chain & Decision Process */ wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy pi that maps states or observations to actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a network of conditional probability, we have * <math> s_1 -> o_1 - (pi_theta) -> a_1 </math> (policy) * <math> s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 </math> (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. We often use and tune hyperparameters for reward functions to make model training faster === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} </math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} </math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let <math> T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k) </math>. r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat Policy Gradients - Directly differentiate objective with respect to the optimal theta and then perform gradient descent Value-based: Estimate value function or q-function of optimal policy (policy is often represented implicitly) Actor-Critic: Estimate value function or q-function of current policy, and find a better policy gradient Model-based: Estimate some transition model, and then use it to improve a policy === REINFORCE === - === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here <math>\alpha</math> represents the learning rate, which is how much new information is weighted relative to old information, while <math>\gamma</math> represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. <math> Q_i+1(s,a)=(1−\alpha)Q_i(s,a)+\alpha(r(s, a)+\gammaV_i(s')) </math> Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. 6982fc6458d9d8573052db52bcf3e21a10dbd8bd 1167 1166 2024-05-21T18:18:08Z Ben 2 /* Q Learning */ wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy pi that maps states or observations to actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a network of conditional probability, we have * <math> s_1 -> o_1 - (pi_theta) -> a_1 </math> (policy) * <math> s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 </math> (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. We often use and tune hyperparameters for reward functions to make model training faster === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} </math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} </math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let <math> T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k) </math>. r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat Policy Gradients - Directly differentiate objective with respect to the optimal theta and then perform gradient descent Value-based: Estimate value function or q-function of optimal policy (policy is often represented implicitly) Actor-Critic: Estimate value function or q-function of current policy, and find a better policy gradient Model-based: Estimate some transition model, and then use it to improve a policy === REINFORCE === - === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here <math>\alpha</math> represents the learning rate, which is how much new information is weighted relative to old information, while <math>\gamma</math> represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. <math> Q_{i+1} (s,a) = (1 - \alpha) Q_i (s,a) + \alpha (r(s,a) + \gamma V_i(s'))</math> Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. c910359bfda75bbdd24c2678f6998b20c330dd71 0 257 1160 1116 2024-05-21T05:21:02Z Vrtnis 21 /* Build the mujoco_wasm Binary */ wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == First git clone <code> https://github.com/zalo/mujoco_wasm </code> Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Running in Browser == Run this in your mujoco folder to start a server. <code> python -m http.server 8000 </code> Then navigate to: <code> http://localhost:8000/index.html </code> [[File:Wasm screenshot13-40-40.png|800px|thumb|none|MuJoCo running in browser]] 5ff174df1e8eb83e2912c817c31370188123fb8d 0 271 1165 2024-05-21T18:11:24Z Ben 2 Created page with "Allen's algorithm learning notes" wikitext text/x-wiki Allen's algorithm learning notes 8218414d6d987f94b48002efb7d9fb69d504b2db 1168 1165 2024-05-21T18:18:47Z Ben 2 wikitext text/x-wiki Isaac's algorithm learning notes 1a36deceabfaf2d0a46ddc0c6a90bd98804db217 1177 1168 2024-05-22T00:11:46Z Is2ac 30 wikitext text/x-wiki Isaac's Beam Search learning notes === Links === * [https://www.width.ai/post/what-is-beam-search#:~:text=Beam%20search%20is%20an%20algorithm,probability%20or%20next%20output%20character. Width.ai Beam Search] * [https://towardsdatascience.com/foundations-of-nlp-explained-visually-beam-search-how-it-works-1586b9849a24 towarddatascience.com visuals] *[https://en.wikipedia.org/wiki/Log_probability Log probability wiki] *[https://en.wikipedia.org/wiki/Softmax_function Softmax_function wiki] [[Category:Beam Search]] === Motivation === Beam Search is used in the context of AI as a final decision making layer for many NLP and speech recognition models. Example: Sequence to Sequence (Seq2Seq) Modeling Task: Translate a sentence from Arabic to English. Overview: * '''1. Tokenization/Normalization:''' Arabic sentence is split into tokens designed for the LLM and normalized * '''2. Encoding:''' LLM encodes the tokens into numerical tokens and generates a sequence of hidden states to represent the input sentence * '''3. Initialize Beam Search:''' Determine the parameters of Beam Search and the decoder's initial states * '''4. Decoding:''' Begin with start-of-sequence token. Model generates probabilities for the next token in each sequence and passes them through an output layer, including a softmax function that normalizes probabilities. Beam Search chooses which paths to continue following and prunes the rest. * '''5. Finalize Output:''' Select Beam with highest probability as final translation, convert tokenized output back into a readable sentence with punctuation, capitalization, etc. Now, to break down the steps of Beam Search Conditional Probability Notes: * Sequence to Sequence search algorithms are based on '''Conditional Probability''' -> describe the likelihood of an event happening given that another event or sequence of events has already occurred * i.e the probability of a new token given the existing sequence is NOT independent and can be calculated <math>Prob(ABC) = Prob(AB) * Prob(C|AB)</math> where <math>Prob(C|AB)</math> is the probability of <math>C</math> occurring given that <math>AB</math> have already occurred * As a result, the graph we run the algorithm on is NOT a Markov Chain Graph, so states are dependent. === Naive Approach #1: Greedy === * Greedy Search takes the best solution at each state in the graph, regardless of previous leaves or future leaves in the sequence. In the context of sequence to sequence, greedy search takes the highest probability word at each position in the sequence and takes it as part of the output. Let's take <math>A, B, C</math> as possible words for our sequence with initial probabilities to be <math>A_0=0.2, B_0=0.5, C_0=0.3</math> The greedy search will take B, the word with highest probability, to be the first word in our sequence. It will continue as such until the end of the sentence is reached. This greedy strategy may be optimal at the current spot of the sequence, but as one might predict, the algorithm struggles with larger outputs where such a path is not the optimal. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(nm)</math> === Naive Approach #2: Breadth First Search === The BFS approach considers every possible sequence of words and outputs the highest probability sequence among all. For example, let's take <math>A, B, C</math> as possible words for our sequence. (Assuming the length of the sequence is fixed at 3) BFS will find the maximum of <math>Prob(AAA), Prob(AAB), ... Prob(CCC)</math> and output that respective sequence BFS is guaranteed to find the optimal sequence, but its runtime is far too large to be feasible for use. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(n^m)</math> === Beam Search === Beam Search is a heuristic algorithm which combines the ideas of the Greedy and BFS algorithms. * Maintains a 'beam width' amount of the most promising sequences * At each iteration add all possible continuations of candidate sequences, and takes the 'beam width' best from those to repeat the algorithm * After the algorithm is finished we take the best option from the candidate sequences as our output If we have at most <math>m</math> words in our final sentence, each state has <math>n</math> word options, and we maintain a maximum bin width size of <math>k</math>, this algorithm runs with a time complexity of <math>O(nmk log(k)</math> [[File:Screenshot 2024-05-21 at 7.50.48 PM.png|thumb]] '''Top-K Sampling for Beam Search''': *<math>k</math> -> maximum width of beams during search *all other probabilities are set to 0 *allows for a more consistent runtime * lower value => faster runtime, less diversity, less optimal results '''Top-P Sampling for Beam Search (Nucleus Sampling)''': *<math>p</math> -> maximum sum of probabilities among all sequences *usually dynamically determined based on the cumulative probabilities of the tokens in the probability distribution -> ensures a certain proportion of the probability mass is considered *top-p sampling allows for more diversity in generated sequences * lower value => faster runtime, less diversity, less optimal results === Other Notes ==== * Softmax Function: [[File:Screenshot 2024-05-21 at 6.56.45 PM.png|thumb]] ensures that all probabilities at each state are in the range (0,1) and sum to 1 * Using semi-log plot to deal with very small probability values trick: Given <math>P</math> as the cumulative probability score of all words in the sequence so far, <math>P=p_0*p_1*p_2...p_n</math> * Since the value of <math>P</math> can become very small, we can run into computational rounding errors. One strategy around this is to take the natural log of the summation of all p values and use this to compare values of <math>P</math>. This works because * <math>P=p_0*p_1*p_2...p_n</math> * <math>log(P)=log(p_0)+log(p_1)+log(p_2)...+log(p_n)</math> * if <math>(P_1>P_2) => log(P_1)>log(P_2)</math>, so values can still be compared in this form c268f8230d0bf604704132732db965dae90cbc44 1178 1177 2024-05-22T00:12:18Z Is2ac 30 /* Beam Search */ wikitext text/x-wiki Isaac's Beam Search learning notes === Links === * [https://www.width.ai/post/what-is-beam-search#:~:text=Beam%20search%20is%20an%20algorithm,probability%20or%20next%20output%20character. Width.ai Beam Search] * [https://towardsdatascience.com/foundations-of-nlp-explained-visually-beam-search-how-it-works-1586b9849a24 towarddatascience.com visuals] *[https://en.wikipedia.org/wiki/Log_probability Log probability wiki] *[https://en.wikipedia.org/wiki/Softmax_function Softmax_function wiki] [[Category:Beam Search]] === Motivation === Beam Search is used in the context of AI as a final decision making layer for many NLP and speech recognition models. Example: Sequence to Sequence (Seq2Seq) Modeling Task: Translate a sentence from Arabic to English. Overview: * '''1. Tokenization/Normalization:''' Arabic sentence is split into tokens designed for the LLM and normalized * '''2. Encoding:''' LLM encodes the tokens into numerical tokens and generates a sequence of hidden states to represent the input sentence * '''3. Initialize Beam Search:''' Determine the parameters of Beam Search and the decoder's initial states * '''4. Decoding:''' Begin with start-of-sequence token. Model generates probabilities for the next token in each sequence and passes them through an output layer, including a softmax function that normalizes probabilities. Beam Search chooses which paths to continue following and prunes the rest. * '''5. Finalize Output:''' Select Beam with highest probability as final translation, convert tokenized output back into a readable sentence with punctuation, capitalization, etc. Now, to break down the steps of Beam Search Conditional Probability Notes: * Sequence to Sequence search algorithms are based on '''Conditional Probability''' -> describe the likelihood of an event happening given that another event or sequence of events has already occurred * i.e the probability of a new token given the existing sequence is NOT independent and can be calculated <math>Prob(ABC) = Prob(AB) * Prob(C|AB)</math> where <math>Prob(C|AB)</math> is the probability of <math>C</math> occurring given that <math>AB</math> have already occurred * As a result, the graph we run the algorithm on is NOT a Markov Chain Graph, so states are dependent. === Naive Approach #1: Greedy === * Greedy Search takes the best solution at each state in the graph, regardless of previous leaves or future leaves in the sequence. In the context of sequence to sequence, greedy search takes the highest probability word at each position in the sequence and takes it as part of the output. Let's take <math>A, B, C</math> as possible words for our sequence with initial probabilities to be <math>A_0=0.2, B_0=0.5, C_0=0.3</math> The greedy search will take B, the word with highest probability, to be the first word in our sequence. It will continue as such until the end of the sentence is reached. This greedy strategy may be optimal at the current spot of the sequence, but as one might predict, the algorithm struggles with larger outputs where such a path is not the optimal. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(nm)</math> === Naive Approach #2: Breadth First Search === The BFS approach considers every possible sequence of words and outputs the highest probability sequence among all. For example, let's take <math>A, B, C</math> as possible words for our sequence. (Assuming the length of the sequence is fixed at 3) BFS will find the maximum of <math>Prob(AAA), Prob(AAB), ... Prob(CCC)</math> and output that respective sequence BFS is guaranteed to find the optimal sequence, but its runtime is far too large to be feasible for use. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(n^m)</math> === Beam Search === Beam Search is a heuristic algorithm which combines the ideas of the Greedy and BFS algorithms. * Maintains a 'beam width' amount of the most promising sequences * At each iteration add all possible continuations of candidate sequences, and takes the 'beam width' best from those to repeat the algorithm * After the algorithm is finished we take the best option from the candidate sequences as our output If we have at most <math>m</math> words in our final sentence, each state has <math>n</math> word options, and we maintain a maximum bin width size of <math>k</math>, this algorithm runs with a time complexity of <math>O(nmk log(k))</math> [[File:Screenshot 2024-05-21 at 7.50.48 PM.png|thumb]] '''Top-K Sampling for Beam Search''': *<math>k</math> -> maximum width of beams during search *all other probabilities are set to 0 *allows for a more consistent runtime * lower value => faster runtime, less diversity, less optimal results '''Top-P Sampling for Beam Search (Nucleus Sampling)''': *<math>p</math> -> maximum sum of probabilities among all sequences *usually dynamically determined based on the cumulative probabilities of the tokens in the probability distribution -> ensures a certain proportion of the probability mass is considered *top-p sampling allows for more diversity in generated sequences * lower value => faster runtime, less diversity, less optimal results === Other Notes ==== * Softmax Function: [[File:Screenshot 2024-05-21 at 6.56.45 PM.png|thumb]] ensures that all probabilities at each state are in the range (0,1) and sum to 1 * Using semi-log plot to deal with very small probability values trick: Given <math>P</math> as the cumulative probability score of all words in the sequence so far, <math>P=p_0*p_1*p_2...p_n</math> * Since the value of <math>P</math> can become very small, we can run into computational rounding errors. One strategy around this is to take the natural log of the summation of all p values and use this to compare values of <math>P</math>. This works because * <math>P=p_0*p_1*p_2...p_n</math> * <math>log(P)=log(p_0)+log(p_1)+log(p_2)...+log(p_n)</math> * if <math>(P_1>P_2) => log(P_1)>log(P_2)</math>, so values can still be compared in this form 300d0d6bfac1083948620ac4f714c86d836569b1 1179 1178 2024-05-22T00:13:10Z Is2ac 30 /* Other Notes = */ wikitext text/x-wiki Isaac's Beam Search learning notes === Links === * [https://www.width.ai/post/what-is-beam-search#:~:text=Beam%20search%20is%20an%20algorithm,probability%20or%20next%20output%20character. Width.ai Beam Search] * [https://towardsdatascience.com/foundations-of-nlp-explained-visually-beam-search-how-it-works-1586b9849a24 towarddatascience.com visuals] *[https://en.wikipedia.org/wiki/Log_probability Log probability wiki] *[https://en.wikipedia.org/wiki/Softmax_function Softmax_function wiki] [[Category:Beam Search]] === Motivation === Beam Search is used in the context of AI as a final decision making layer for many NLP and speech recognition models. Example: Sequence to Sequence (Seq2Seq) Modeling Task: Translate a sentence from Arabic to English. Overview: * '''1. Tokenization/Normalization:''' Arabic sentence is split into tokens designed for the LLM and normalized * '''2. Encoding:''' LLM encodes the tokens into numerical tokens and generates a sequence of hidden states to represent the input sentence * '''3. Initialize Beam Search:''' Determine the parameters of Beam Search and the decoder's initial states * '''4. Decoding:''' Begin with start-of-sequence token. Model generates probabilities for the next token in each sequence and passes them through an output layer, including a softmax function that normalizes probabilities. Beam Search chooses which paths to continue following and prunes the rest. * '''5. Finalize Output:''' Select Beam with highest probability as final translation, convert tokenized output back into a readable sentence with punctuation, capitalization, etc. Now, to break down the steps of Beam Search Conditional Probability Notes: * Sequence to Sequence search algorithms are based on '''Conditional Probability''' -> describe the likelihood of an event happening given that another event or sequence of events has already occurred * i.e the probability of a new token given the existing sequence is NOT independent and can be calculated <math>Prob(ABC) = Prob(AB) * Prob(C|AB)</math> where <math>Prob(C|AB)</math> is the probability of <math>C</math> occurring given that <math>AB</math> have already occurred * As a result, the graph we run the algorithm on is NOT a Markov Chain Graph, so states are dependent. === Naive Approach #1: Greedy === * Greedy Search takes the best solution at each state in the graph, regardless of previous leaves or future leaves in the sequence. In the context of sequence to sequence, greedy search takes the highest probability word at each position in the sequence and takes it as part of the output. Let's take <math>A, B, C</math> as possible words for our sequence with initial probabilities to be <math>A_0=0.2, B_0=0.5, C_0=0.3</math> The greedy search will take B, the word with highest probability, to be the first word in our sequence. It will continue as such until the end of the sentence is reached. This greedy strategy may be optimal at the current spot of the sequence, but as one might predict, the algorithm struggles with larger outputs where such a path is not the optimal. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(nm)</math> === Naive Approach #2: Breadth First Search === The BFS approach considers every possible sequence of words and outputs the highest probability sequence among all. For example, let's take <math>A, B, C</math> as possible words for our sequence. (Assuming the length of the sequence is fixed at 3) BFS will find the maximum of <math>Prob(AAA), Prob(AAB), ... Prob(CCC)</math> and output that respective sequence BFS is guaranteed to find the optimal sequence, but its runtime is far too large to be feasible for use. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(n^m)</math> === Beam Search === Beam Search is a heuristic algorithm which combines the ideas of the Greedy and BFS algorithms. * Maintains a 'beam width' amount of the most promising sequences * At each iteration add all possible continuations of candidate sequences, and takes the 'beam width' best from those to repeat the algorithm * After the algorithm is finished we take the best option from the candidate sequences as our output If we have at most <math>m</math> words in our final sentence, each state has <math>n</math> word options, and we maintain a maximum bin width size of <math>k</math>, this algorithm runs with a time complexity of <math>O(nmk log(k))</math> [[File:Screenshot 2024-05-21 at 7.50.48 PM.png|thumb]] '''Top-K Sampling for Beam Search''': *<math>k</math> -> maximum width of beams during search *all other probabilities are set to 0 *allows for a more consistent runtime * lower value => faster runtime, less diversity, less optimal results '''Top-P Sampling for Beam Search (Nucleus Sampling)''': *<math>p</math> -> maximum sum of probabilities among all sequences *usually dynamically determined based on the cumulative probabilities of the tokens in the probability distribution -> ensures a certain proportion of the probability mass is considered *top-p sampling allows for more diversity in generated sequences * lower value => faster runtime, less diversity, less optimal results === Other Notes ==== '''Softmax Function:''' [[File:Screenshot 2024-05-21 at 6.56.45 PM.png|thumb]] * ensures that all probabilities at each state are in the range (0,1) and sum to 1 '''Semi-log plot trick''' * Using semi-log plot to deal with very small probability values trick: Given <math>P</math> as the cumulative probability score of all words in the sequence so far, <math>P=p_0*p_1*p_2...p_n</math> * Since the value of <math>P</math> can become very small, we can run into computational rounding errors. One strategy around this is to take the natural log of the summation of all p values and use this to compare values of <math>P</math>. This works because * <math>P=p_0*p_1*p_2...p_n</math> * <math>log(P)=log(p_0)+log(p_1)+log(p_2)...+log(p_n)</math> * if <math>(P_1>P_2) => log(P_1)>log(P_2)</math>, so values can still be compared in this form 1be513bfeb7e40dfcd7c48cd377767f1119db532 1180 1179 2024-05-22T00:40:27Z Is2ac 30 /* Other Notes = */ wikitext text/x-wiki Isaac's Beam Search learning notes === Links === * [https://www.width.ai/post/what-is-beam-search#:~:text=Beam%20search%20is%20an%20algorithm,probability%20or%20next%20output%20character. Width.ai Beam Search] * [https://towardsdatascience.com/foundations-of-nlp-explained-visually-beam-search-how-it-works-1586b9849a24 towarddatascience.com visuals] *[https://en.wikipedia.org/wiki/Log_probability Log probability wiki] *[https://en.wikipedia.org/wiki/Softmax_function Softmax_function wiki] [[Category:Beam Search]] === Motivation === Beam Search is used in the context of AI as a final decision making layer for many NLP and speech recognition models. Example: Sequence to Sequence (Seq2Seq) Modeling Task: Translate a sentence from Arabic to English. Overview: * '''1. Tokenization/Normalization:''' Arabic sentence is split into tokens designed for the LLM and normalized * '''2. Encoding:''' LLM encodes the tokens into numerical tokens and generates a sequence of hidden states to represent the input sentence * '''3. Initialize Beam Search:''' Determine the parameters of Beam Search and the decoder's initial states * '''4. Decoding:''' Begin with start-of-sequence token. Model generates probabilities for the next token in each sequence and passes them through an output layer, including a softmax function that normalizes probabilities. Beam Search chooses which paths to continue following and prunes the rest. * '''5. Finalize Output:''' Select Beam with highest probability as final translation, convert tokenized output back into a readable sentence with punctuation, capitalization, etc. Now, to break down the steps of Beam Search Conditional Probability Notes: * Sequence to Sequence search algorithms are based on '''Conditional Probability''' -> describe the likelihood of an event happening given that another event or sequence of events has already occurred * i.e the probability of a new token given the existing sequence is NOT independent and can be calculated <math>Prob(ABC) = Prob(AB) * Prob(C|AB)</math> where <math>Prob(C|AB)</math> is the probability of <math>C</math> occurring given that <math>AB</math> have already occurred * As a result, the graph we run the algorithm on is NOT a Markov Chain Graph, so states are dependent. === Naive Approach #1: Greedy === * Greedy Search takes the best solution at each state in the graph, regardless of previous leaves or future leaves in the sequence. In the context of sequence to sequence, greedy search takes the highest probability word at each position in the sequence and takes it as part of the output. Let's take <math>A, B, C</math> as possible words for our sequence with initial probabilities to be <math>A_0=0.2, B_0=0.5, C_0=0.3</math> The greedy search will take B, the word with highest probability, to be the first word in our sequence. It will continue as such until the end of the sentence is reached. This greedy strategy may be optimal at the current spot of the sequence, but as one might predict, the algorithm struggles with larger outputs where such a path is not the optimal. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(nm)</math> === Naive Approach #2: Breadth First Search === The BFS approach considers every possible sequence of words and outputs the highest probability sequence among all. For example, let's take <math>A, B, C</math> as possible words for our sequence. (Assuming the length of the sequence is fixed at 3) BFS will find the maximum of <math>Prob(AAA), Prob(AAB), ... Prob(CCC)</math> and output that respective sequence BFS is guaranteed to find the optimal sequence, but its runtime is far too large to be feasible for use. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(n^m)</math> === Beam Search === Beam Search is a heuristic algorithm which combines the ideas of the Greedy and BFS algorithms. * Maintains a 'beam width' amount of the most promising sequences * At each iteration add all possible continuations of candidate sequences, and takes the 'beam width' best from those to repeat the algorithm * After the algorithm is finished we take the best option from the candidate sequences as our output If we have at most <math>m</math> words in our final sentence, each state has <math>n</math> word options, and we maintain a maximum bin width size of <math>k</math>, this algorithm runs with a time complexity of <math>O(nmk log(k))</math> [[File:Screenshot 2024-05-21 at 7.50.48 PM.png|thumb]] '''Top-K Sampling for Beam Search''': *<math>k</math> -> maximum width of beams during search *all other probabilities are set to 0 *allows for a more consistent runtime * lower value => faster runtime, less diversity, less optimal results '''Top-P Sampling for Beam Search (Nucleus Sampling)''': *<math>p</math> -> maximum sum of probabilities among all sequences *usually dynamically determined based on the cumulative probabilities of the tokens in the probability distribution -> ensures a certain proportion of the probability mass is considered *top-p sampling allows for more diversity in generated sequences * lower value => faster runtime, less diversity, less optimal results === Other Notes === '''Softmax Function:''' [[File:Screenshot 2024-05-21 at 6.56.45 PM.png|thumb]] * ensures that all probabilities at each state are in the range (0,1) and sum to 1 '''Semi-log plot trick''' * Using semi-log plot to deal with very small probability values trick: Given <math>P</math> as the cumulative probability score of all words in the sequence so far, <math>P=p_0*p_1*p_2...p_n</math> * Since the value of <math>P</math> can become very small, we can run into computational rounding errors. One strategy around this is to take the natural log of the summation of all p values and use this to compare values of <math>P</math>. This works because * <math>P=p_0*p_1*p_2...p_n</math> * <math>log(P)=log(p_0)+log(p_1)+log(p_2)...+log(p_n)</math> * if <math>(P_1>P_2) => log(P_1)>log(P_2)</math>, so values can still be compared in this form f199c71bfedd597637d2e32f2e1b91030186a5d5 1181 1180 2024-05-22T00:47:52Z Is2ac 30 wikitext text/x-wiki Isaac's Beam Search learning notes === Links === * [https://www.width.ai/post/what-is-beam-search#:~:text=Beam%20search%20is%20an%20algorithm,probability%20or%20next%20output%20character. Width.ai Beam Search] * [https://towardsdatascience.com/foundations-of-nlp-explained-visually-beam-search-how-it-works-1586b9849a24 towarddatascience.com visuals] *[https://en.wikipedia.org/wiki/Log_probability Log probability wiki] *[https://en.wikipedia.org/wiki/Softmax_function Softmax_function wiki] [[Category:Beam Search]] === Motivation === Beam Search is used in the context of AI as a final decision making layer for many NLP and speech recognition models. Example: Sequence to Sequence (Seq2Seq) Modeling Task: Translate a sentence from Arabic to English. Overview: * '''1. Tokenization/Normalization:''' Arabic sentence is split into tokens designed for the LLM and normalized * '''2. Encoding:''' LLM encodes the tokens into numerical tokens and generates a sequence of hidden states to represent the input sentence * '''3. Initialize Beam Search:''' Determine the parameters of Beam Search and the decoder's initial states * '''4. Decoding:''' Begin with start-of-sequence token. Model generates probabilities for the next token in each sequence and passes them through an output layer, including a softmax function that normalizes probabilities. Beam Search chooses which paths to continue following and prunes the rest. * '''5. Finalize Output:''' Select Beam with highest probability as final translation, convert tokenized output back into a readable sentence with punctuation, capitalization, etc. Conditional Probability Notes: * Sequence to Sequence search algorithms are based on '''Conditional Probability''' -> describe the likelihood of an event happening given that another event or sequence of events has already occurred * i.e the probability of a new token given the existing sequence is NOT independent and can be calculated <math>Prob(ABC) = Prob(AB) * Prob(C|AB)</math> where <math>Prob(C|AB)</math> is the probability of <math>C</math> occurring given that <math>AB</math> have already occurred * As a result, the graph we run the algorithm on is NOT a Markov Chain Graph, so states are dependent. === Naive Approach #1: Greedy === * Greedy Search takes the best solution at each state in the graph, regardless of previous leaves or future leaves in the sequence. In the context of sequence to sequence, greedy search takes the highest probability word at each position in the sequence and takes it as part of the output. Let's take <math>A, B, C</math> as possible words for our sequence with initial probabilities to be <math>A_0=0.2, B_0=0.5, C_0=0.3</math> The greedy search will take B, the word with highest probability, to be the first word in our sequence. It will continue as such until the end of the sentence is reached. This greedy strategy may be optimal at the current spot of the sequence, but as one might predict, the algorithm struggles with larger outputs where such a path is not the optimal. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(nm)</math> === Naive Approach #2: Breadth First Search === The BFS approach considers every possible sequence of words and outputs the highest probability sequence among all. For example, let's take <math>A, B, C</math> as possible words for our sequence. (Assuming the length of the sequence is fixed at 3) BFS will find the maximum of <math>Prob(AAA), Prob(AAB), ... Prob(CCC)</math> and output that respective sequence BFS is guaranteed to find the optimal sequence, but its runtime is far too large to be feasible for use. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(n^m)</math> === Beam Search === Beam Search is a heuristic algorithm which combines the ideas of the Greedy and BFS algorithms. * Maintains a 'beam width' amount of the most promising sequences * At each iteration add all possible continuations of candidate sequences, and takes the 'beam width' best from those to repeat the algorithm * After the algorithm is finished we take the best option from the candidate sequences as our output If we have at most <math>m</math> words in our final sentence, each state has <math>n</math> word options, and we maintain a maximum bin width size of <math>k</math>, this algorithm runs with a time complexity of <math>O(nmk log(k))</math> [[File:Screenshot 2024-05-21 at 7.50.48 PM.png|thumb]] '''Top-K Sampling for Beam Search''': *<math>k</math> -> maximum width of beams during search *all other probabilities are set to 0 *allows for a more consistent runtime * lower value => faster runtime, less diversity, less optimal results '''Top-P Sampling for Beam Search (Nucleus Sampling)''': *<math>p</math> -> maximum sum of probabilities among all sequences *usually dynamically determined based on the cumulative probabilities of the tokens in the probability distribution -> ensures a certain proportion of the probability mass is considered *top-p sampling allows for more diversity in generated sequences * lower value => faster runtime, less diversity, less optimal results === Other Notes === '''Softmax Function:''' [[File:Screenshot 2024-05-21 at 6.56.45 PM.png|thumb]] * ensures that all probabilities at each state are in the range (0,1) and sum to 1 '''Semi-log plot trick''' * Using semi-log plot to deal with very small probability values trick: Given <math>P</math> as the cumulative probability score of all words in the sequence so far, <math>P=p_0*p_1*p_2...p_n</math> * Since the value of <math>P</math> can become very small, we can run into computational rounding errors. One strategy around this is to take the natural log of the summation of all p values and use this to compare values of <math>P</math>. This works because * <math>P=p_0*p_1*p_2...p_n</math> * <math>log(P)=log(p_0)+log(p_1)+log(p_2)...+log(p_n)</math> * if <math>(P_1>P_2) => log(P_1)>log(P_2)</math>, so values can still be compared in this form c9ce6d4d967beefa8025225be184a4442b6adb7b 1182 1181 2024-05-22T00:55:02Z Ben 2 /* Other Notes */ wikitext text/x-wiki Isaac's Beam Search learning notes === Links === * [https://www.width.ai/post/what-is-beam-search#:~:text=Beam%20search%20is%20an%20algorithm,probability%20or%20next%20output%20character. Width.ai Beam Search] * [https://towardsdatascience.com/foundations-of-nlp-explained-visually-beam-search-how-it-works-1586b9849a24 towarddatascience.com visuals] *[https://en.wikipedia.org/wiki/Log_probability Log probability wiki] *[https://en.wikipedia.org/wiki/Softmax_function Softmax_function wiki] [[Category:Beam Search]] === Motivation === Beam Search is used in the context of AI as a final decision making layer for many NLP and speech recognition models. Example: Sequence to Sequence (Seq2Seq) Modeling Task: Translate a sentence from Arabic to English. Overview: * '''1. Tokenization/Normalization:''' Arabic sentence is split into tokens designed for the LLM and normalized * '''2. Encoding:''' LLM encodes the tokens into numerical tokens and generates a sequence of hidden states to represent the input sentence * '''3. Initialize Beam Search:''' Determine the parameters of Beam Search and the decoder's initial states * '''4. Decoding:''' Begin with start-of-sequence token. Model generates probabilities for the next token in each sequence and passes them through an output layer, including a softmax function that normalizes probabilities. Beam Search chooses which paths to continue following and prunes the rest. * '''5. Finalize Output:''' Select Beam with highest probability as final translation, convert tokenized output back into a readable sentence with punctuation, capitalization, etc. Conditional Probability Notes: * Sequence to Sequence search algorithms are based on '''Conditional Probability''' -> describe the likelihood of an event happening given that another event or sequence of events has already occurred * i.e the probability of a new token given the existing sequence is NOT independent and can be calculated <math>Prob(ABC) = Prob(AB) * Prob(C|AB)</math> where <math>Prob(C|AB)</math> is the probability of <math>C</math> occurring given that <math>AB</math> have already occurred * As a result, the graph we run the algorithm on is NOT a Markov Chain Graph, so states are dependent. === Naive Approach #1: Greedy === * Greedy Search takes the best solution at each state in the graph, regardless of previous leaves or future leaves in the sequence. In the context of sequence to sequence, greedy search takes the highest probability word at each position in the sequence and takes it as part of the output. Let's take <math>A, B, C</math> as possible words for our sequence with initial probabilities to be <math>A_0=0.2, B_0=0.5, C_0=0.3</math> The greedy search will take B, the word with highest probability, to be the first word in our sequence. It will continue as such until the end of the sentence is reached. This greedy strategy may be optimal at the current spot of the sequence, but as one might predict, the algorithm struggles with larger outputs where such a path is not the optimal. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(nm)</math> === Naive Approach #2: Breadth First Search === The BFS approach considers every possible sequence of words and outputs the highest probability sequence among all. For example, let's take <math>A, B, C</math> as possible words for our sequence. (Assuming the length of the sequence is fixed at 3) BFS will find the maximum of <math>Prob(AAA), Prob(AAB), ... Prob(CCC)</math> and output that respective sequence BFS is guaranteed to find the optimal sequence, but its runtime is far too large to be feasible for use. If we have at most <math>m</math> words in our final sentence and each state has <math>n</math> word options, this algorithm runs with a time complexity of <math>O(n^m)</math> === Beam Search === Beam Search is a heuristic algorithm which combines the ideas of the Greedy and BFS algorithms. * Maintains a 'beam width' amount of the most promising sequences * At each iteration add all possible continuations of candidate sequences, and takes the 'beam width' best from those to repeat the algorithm * After the algorithm is finished we take the best option from the candidate sequences as our output If we have at most <math>m</math> words in our final sentence, each state has <math>n</math> word options, and we maintain a maximum bin width size of <math>k</math>, this algorithm runs with a time complexity of <math>O(nmk log(k))</math> [[File:Screenshot 2024-05-21 at 7.50.48 PM.png|thumb]] '''Top-K Sampling for Beam Search''': *<math>k</math> -> maximum width of beams during search *all other probabilities are set to 0 *allows for a more consistent runtime * lower value => faster runtime, less diversity, less optimal results '''Top-P Sampling for Beam Search (Nucleus Sampling)''': *<math>p</math> -> maximum sum of probabilities among all sequences *usually dynamically determined based on the cumulative probabilities of the tokens in the probability distribution -> ensures a certain proportion of the probability mass is considered *top-p sampling allows for more diversity in generated sequences * lower value => faster runtime, less diversity, less optimal results === Other Notes === '''Softmax Function:''' [[File:Screenshot 2024-05-21 at 6.56.45 PM.png|thumb]] * ensures that all probabilities at each state are in the range (0,1) and sum to 1 '''Log semiring trick''' * Using [https://en.wikipedia.org/wiki/Log_semiring log semiring] to deal with very small probability values trick: Given <math>P</math> as the cumulative probability score of all words in the sequence so far, <math>P=p_0*p_1*p_2...p_n</math> * Since the value of <math>P</math> can become very small, we can run into computational rounding errors. One strategy around this is to take the natural log of the summation of all p values and use this to compare values of <math>P</math>. This works because * <math>P=p_0*p_1*p_2...p_n</math> * <math>log(P)=log(p_0)+log(p_1)+log(p_2)...+log(p_n)</math> * if <math>(P_1>P_2) => log(P_1)>log(P_2)</math>, so values can still be compared in this form 8ced1c13c9ceff0d36270f3a91f3d79e29144d33 0 265 1169 1132 2024-05-21T18:27:46Z Ben 2 /* Extras */ wikitext text/x-wiki This document contains a walk-through of how to set up MediaWiki on AWS. In total, this process should take about 30 minutes. === Getting Started === # Use [https://aws.amazon.com/marketplace/pp/prodview-3tokjpxwvddp2 this service] to install MediaWiki to an EC2 instance # After installing, SSH into the instance. The MediaWiki root files live in <code>/var/www/html</code>, and Apache configuration files live in <code>/etc/apache2</code> === Set up A Record === In your website DNS, create an A record which points from your desired URL to the IP address of the newly-created EC2 instance. For example: * '''Host name''': <code>@</code> (for the main domains) or <code>wiki</code> (for subdomains) * '''Type''': <code>A</code> * '''Data''': <code>127.0.0.1</code> (use the IP address of the EC2 instance) === Installing OpenSSL === Roughly follow [https://www.digitalocean.com/community/tutorials/how-to-secure-apache-with-let-s-encrypt-on-ubuntu-20-04 these instructions] to install OpenSSL using LetsEncrypt # Install certbot <syntaxhighlight lang="bash"> sudo apt install certbot python3-certbot-apache </syntaxhighlight> # Run certbot (you can just select "No redirect" <syntaxhighlight lang="bash"> sudo certbot --apache </syntaxhighlight> # Verify certbot <syntaxhighlight lang="bash"> sudo systemctl status certbot.timer </syntaxhighlight> === Update Page URLs === To get MediaWiki to display pages using the <code>/w/Some_Page</code> URL, modify <code>LocalSettings.php</code>. Replace the line where it says <code>$wgScriptPage = "";</code> with the following: <syntaxhighlight lang="php"> $wgScriptPath = ""; $wgArticlePath = "/w/$1"; $wgUsePathInfo = true; </syntaxhighlight> Next, in <code>/etc/apache2/apache2.conf</code>, add the following section: <syntaxhighlight lang="text"> <Directory /var/www/html> AllowOverride all </Directory> </syntaxhighlight> Create a new <code>.htaccess</code> in <code>/var/www/html</code> with the following contents: <syntaxhighlight lang="text"> RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule ^w/(.*)$ index.php/$1 [PT,L,QSA] </syntaxhighlight> Finally, reload Apache2: <syntaxhighlight lang="bash"> sudo service apache2 reload </syntaxhighlight> === Extras === Here are some notes about some extra features you can add on top of MediaWiki. ==== Storing images in S3 with Cloudfront ==== You can use the [https://www.mediawiki.org/wiki/Extension:AWS AWS MediaWiki Extension] to allow users to upload images and files, store them in S3, and serve them through CloudFront. This process is somewhat involved. I may write notes about how to do this well in the future. ==== Enabling LaTeX ==== # Install the [https://www.mediawiki.org/wiki/Extension:Math Math] extension to the <code>extensions</code> directory # Change to the <code>origin/REL1_31</code> branch # Run the update script: <code>sudo php maintenance/update.php</code> 6a48a4427e157ca0e0dc98888d971b4a6221f2f8 1170 1169 2024-05-21T18:31:42Z Ben 2 /* Enabling LaTeX */ wikitext text/x-wiki This document contains a walk-through of how to set up MediaWiki on AWS. In total, this process should take about 30 minutes. === Getting Started === # Use [https://aws.amazon.com/marketplace/pp/prodview-3tokjpxwvddp2 this service] to install MediaWiki to an EC2 instance # After installing, SSH into the instance. The MediaWiki root files live in <code>/var/www/html</code>, and Apache configuration files live in <code>/etc/apache2</code> === Set up A Record === In your website DNS, create an A record which points from your desired URL to the IP address of the newly-created EC2 instance. For example: * '''Host name''': <code>@</code> (for the main domains) or <code>wiki</code> (for subdomains) * '''Type''': <code>A</code> * '''Data''': <code>127.0.0.1</code> (use the IP address of the EC2 instance) === Installing OpenSSL === Roughly follow [https://www.digitalocean.com/community/tutorials/how-to-secure-apache-with-let-s-encrypt-on-ubuntu-20-04 these instructions] to install OpenSSL using LetsEncrypt # Install certbot <syntaxhighlight lang="bash"> sudo apt install certbot python3-certbot-apache </syntaxhighlight> # Run certbot (you can just select "No redirect" <syntaxhighlight lang="bash"> sudo certbot --apache </syntaxhighlight> # Verify certbot <syntaxhighlight lang="bash"> sudo systemctl status certbot.timer </syntaxhighlight> === Update Page URLs === To get MediaWiki to display pages using the <code>/w/Some_Page</code> URL, modify <code>LocalSettings.php</code>. Replace the line where it says <code>$wgScriptPage = "";</code> with the following: <syntaxhighlight lang="php"> $wgScriptPath = ""; $wgArticlePath = "/w/$1"; $wgUsePathInfo = true; </syntaxhighlight> Next, in <code>/etc/apache2/apache2.conf</code>, add the following section: <syntaxhighlight lang="text"> <Directory /var/www/html> AllowOverride all </Directory> </syntaxhighlight> Create a new <code>.htaccess</code> in <code>/var/www/html</code> with the following contents: <syntaxhighlight lang="text"> RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule ^w/(.*)$ index.php/$1 [PT,L,QSA] </syntaxhighlight> Finally, reload Apache2: <syntaxhighlight lang="bash"> sudo service apache2 reload </syntaxhighlight> === Extras === Here are some notes about some extra features you can add on top of MediaWiki. ==== Storing images in S3 with Cloudfront ==== You can use the [https://www.mediawiki.org/wiki/Extension:AWS AWS MediaWiki Extension] to allow users to upload images and files, store them in S3, and serve them through CloudFront. This process is somewhat involved. I may write notes about how to do this well in the future. ==== Enabling LaTeX ==== * Install the <code>REL1_31</code> branch of the [https://www.mediawiki.org/wiki/Extension:Math Math] extension to the <code>extensions</code> directory <syntaxhighlight lang="bash"> sudo git clone -b REL1_31 --single-branch --depth 1 https://gerrit.wikimedia.org/r/mediawiki/extensions/Math </syntaxhighlight> * Enable the extension in <code>LocalSettings.php</code> <syntaxhighlight lang="php"> wfLoadExtension('Math'); </syntaxhighlight> * Run the update script <syntaxhighlight lang="bash"> sudo php maintenance/update.php </syntaxhighlight> cdf268539a8538ebe7b8e58c8d2eb4c83fef9b59 0 272 1171 2024-05-21T21:34:04Z Vrtnis 21 Created page with "== Overview == '''MuJoCo XLA (MJX)''' is a specialized extension of the MuJoCo physics engine, designed to run simulations on hardware supported by the XLA (Accelerated Linea..." wikitext text/x-wiki == Overview == '''MuJoCo XLA (MJX)''' is a specialized extension of the MuJoCo physics engine, designed to run simulations on hardware supported by the XLA (Accelerated Linear Algebra) compiler via the JAX framework. 5a7feddccaac5a741ad21783986ac6d7e1ac7668 1172 1171 2024-05-21T21:34:27Z Vrtnis 21 wikitext text/x-wiki == Overview == '''MuJoCo XLA (MJX)''' is a specialized extension of the [[MuJoCo]] physics engine, designed to run simulations on hardware supported by the XLA (Accelerated Linear Algebra) compiler via the JAX framework. 433afe4e8c0638664314dea0e02964043d76bafa 1173 1172 2024-05-21T21:41:29Z Vrtnis 21 wikitext text/x-wiki == Overview == '''MuJoCo XLA (MJX)''' is a specialized extension of the [[MuJoCo]] physics engine, designed to run simulations on hardware supported by the XLA (Accelerated Linear Algebra) compiler via the JAX framework. === Installation === Install using: <syntaxhighlight lang="bash"> pip install mujoco-mjx </syntaxhighlight> a09ba0fc598c58c1ce59a85894d65d3b99096b20 1174 1173 2024-05-21T21:42:14Z Vrtnis 21 wikitext text/x-wiki == Overview == '''MuJoCo XLA (MJX)''' is a specialized extension of the [[MuJoCo]] physics engine, designed to run simulations on hardware supported by the XLA (Accelerated Linear Algebra) compiler via the JAX framework. === Installation === Install using: <syntaxhighlight lang="bash"> pip install mujoco-mjx </syntaxhighlight> == Colab Tutorial == A detailed tutorial demonstrating the use of MJX along with reinforcement learning to train humanoid and quadruped robots to locomote is available [https://colab.research.google.com/github/google-deepmind/mujoco/blob/main/mjx/tutorial.ipynb#scrollTo=MpkYHwCqk7W- here]. 13b5abfa3817ffe2647e93b862535292b577f8cb 1183 1174 2024-05-22T04:29:30Z Vrtnis 21 wikitext text/x-wiki == Overview == '''MuJoCo XLA (MJX)''' is a specialized extension of the [[MuJoCo]] physics engine, designed to run simulations on hardware supported by the XLA (Accelerated Linear Algebra) compiler via the JAX framework. Running single threaded physics simulation on the GPU is not very efficient. The advantage with MJX is that we can run environments in parallel on a hardware accelerated device. === Installation === Install using: <syntaxhighlight lang="bash"> pip install mujoco-mjx </syntaxhighlight> == Colab Tutorial == A detailed tutorial demonstrating the use of MJX along with reinforcement learning to train humanoid and quadruped robots to locomote is available [https://colab.research.google.com/github/google-deepmind/mujoco/blob/main/mjx/tutorial.ipynb#scrollTo=MpkYHwCqk7W- here]. 4b112ba4d1c8f46e0c74006aedeff361af99c891 6 273 1175 2024-05-21T22:57:36Z Is2ac 30 wikitext text/x-wiki Softmax Formula 64ee9f66ab39e3890d24cb48ad79b824588d7eda 6 274 1176 2024-05-21T23:51:40Z Is2ac 30 wikitext text/x-wiki Beam Search graphic from towardsdatascience.com 177914ee2d51b1c90ff554e31d1b9db386a83d86 0 275 1184 2024-05-22T17:06:46Z Vrtnis 21 Created page with "Humanoid-Gym is an advanced reinforcement learning (RL) framework built on Nvidia Isaac Gym, designed for training locomotion skills in humanoid robots. Notably, it emphasizes..." wikitext text/x-wiki Humanoid-Gym is an advanced reinforcement learning (RL) framework built on Nvidia Isaac Gym, designed for training locomotion skills in humanoid robots. Notably, it emphasizes zero-shot transfer, enabling skills learned in simulation to be directly applied to real-world environments without additional adjustments. 51a7d732d608fe17a24ad34da4891d702f1fa5fd 1185 1184 2024-05-22T17:12:31Z Vrtnis 21 wikitext text/x-wiki Humanoid-Gym is an advanced reinforcement learning (RL) framework built on Nvidia Isaac Gym, designed for training locomotion skills in humanoid robots. Notably, it emphasizes zero-shot transfer, enabling skills learned in simulation to be directly applied to real-world environments without additional adjustments. Humanoid-Gym streamlines the process of training humanoid robots by providing an intuitive and efficient platform. By integrating Nvidia Isaac Gym with MuJoCo, it allows users to test and verify trained policies in various simulation environments. This capability ensures the robustness and versatility of the trained behaviors, facilitating a seamless transition from virtual training to real-world application. e002f792f4a6ff4d57046537591bc929bd8568ab 1186 1185 2024-05-22T17:15:56Z Vrtnis 21 wikitext text/x-wiki Humanoid-Gym is an advanced reinforcement learning (RL) framework built on Nvidia Isaac Gym, designed for training locomotion skills in humanoid robots. Notably, it emphasizes zero-shot transfer, enabling skills learned in simulation to be directly applied to real-world environments without additional adjustments. Humanoid-Gym streamlines the process of training humanoid robots by providing an intuitive and efficient platform. By integrating Nvidia Isaac Gym with MuJoCo, it allows users to test and verify trained policies in various simulation environments. This capability ensures the robustness and versatility of the trained behaviors, facilitating a seamless transition from virtual training to real-world application. [https://github.com/roboterax/humanoid-gym GitHub] fcf22a9520a67b55c8d592e1f05b85c4e54d4ceb 0 276 1187 2024-05-22T17:56:36Z Vrtnis 21 Created page with "K-Scale has an updated fork of [[Humanoid Gym]] available at . One of the significant changes involved modifying how the framework handles the initialization of simulation d..." wikitext text/x-wiki K-Scale has an updated fork of [[Humanoid Gym]] available at . One of the significant changes involved modifying how the framework handles the initialization of simulation data. Previously, the framework used a fixed dimension for reshaping tensors, y. a60274c3bc3057f2793dcd50eb5aa75ab435ce97 1188 1187 2024-05-22T18:03:32Z Vrtnis 21 /* Adding framework changes */ wikitext text/x-wiki K-Scale has an updated fork of [[Humanoid Gym]] available at . One of the significant changes involved modifying how the framework handles the initialization of simulation data. Previously, the framework used a fixed dimension for reshaping tensors, which limited its flexibility in handling different numbers of bodies in the simulation. The recent update adjusted this process to allow for any number of bodies, thereby improving the framework's ability to manage various simulation scenarios more effectively. 8e041ab68f69f01f1caff3bd9cb1bc668131d120 1189 1188 2024-05-22T18:04:49Z Vrtnis 21 /*add gh url*/ wikitext text/x-wiki K-Scale has an updated fork of [[Humanoid Gym]] available at this [https://github.com/kscalelabs/humanoid-gym GitHub repository]. One of the significant changes involved modifying how the framework handles the initialization of simulation data. Previously, the framework used a fixed dimension for reshaping tensors, which limited its flexibility in handling different numbers of bodies in the simulation. The recent update adjusted this process to allow for any number of bodies, thereby improving the framework's ability to manage various simulation scenarios more effectively. f12efa63ab171433d61456e839061a87e825c85d 1190 1189 2024-05-22T18:06:47Z Vrtnis 21 /*added graphics handling changes*/ wikitext text/x-wiki K-Scale has an updated fork of [[Humanoid Gym]] available at this [https://github.com/kscalelabs/humanoid-gym GitHub repository]. One of the significant changes involved modifying how the framework handles the initialization of simulation data. Previously, the framework used a fixed dimension for reshaping tensors, which limited its flexibility in handling different numbers of bodies in the simulation. The recent update adjusted this process to allow for any number of bodies, thereby improving the framework's ability to manage various simulation scenarios more effectively. Another important enhancement was the optimization of graphics device handling, especially in headless mode where rendering is not required. The update introduced a conditional setting that disables the graphics device when the simulation runs in headless mode. This change helps improve performance by avoiding unnecessary rendering operations, making the framework more efficient for large-scale simulations where visual output is not needed. 1072d89d4a8fa9df0c78ff569d4e97b725b56c03 1191 1190 2024-05-22T18:12:54Z Vrtnis 21 /* add descriptive titles*/ wikitext text/x-wiki K-Scale has an updated fork of [[Humanoid Gym]] available at this [https://github.com/kscalelabs/humanoid-gym GitHub repository]. ==== Updates to initialization of simulation data ==== One of the significant changes involved modifying how the framework handles the initialization of simulation data. Previously, the framework used a fixed dimension for reshaping tensors, which limited its flexibility in handling different numbers of bodies in the simulation. The recent update adjusted this process to allow for any number of bodies, thereby improving the framework's ability to manage various simulation scenarios more effectively. ==== Optimization of graphics device handling ==== Another important enhancement was the optimization of graphics device handling, especially in headless mode where rendering is not required. The update introduced a conditional setting that disables the graphics device when the simulation runs in headless mode. This change helps improve performance by avoiding unnecessary rendering operations, making the framework more efficient for large-scale simulations where visual output is not needed. 43625a6a5e8fb5c4fb0320323c20c4f93360e7be 0 277 1192 2024-05-22T18:56:55Z Vrtnis 21 Created page with "= K-Scale Sim Library = A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface fo..." wikitext text/x-wiki = K-Scale Sim Library = A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. 6549b764457d43567a2d448eb3a16338c6d314c8 1193 1192 2024-05-22T18:57:15Z Vrtnis 21 /* K-Scale Sim Library */ wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. 0b8dee7fb0c3b4ceb5d0284ea3bdf6da1021bc13 1194 1193 2024-05-22T18:58:20Z Vrtnis 21 wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! 07d9bcfed2a7adad55f2e78e20d09a1afa17ccc2 1195 1194 2024-05-22T18:59:03Z Vrtnis 21 wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. afcd3ac099764ed2fa5792523672d7a0bcc2aca2 1196 1195 2024-05-22T19:00:26Z Vrtnis 21 wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> df8d2e96351e6e1b1cfa3705f4f8859d1a35f90b 1197 1196 2024-05-22T19:01:09Z Vrtnis 21 wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> Install third-party dependencies: Manually download IsaacGym_Preview_4_Package.tar.gz from [https://developer.nvidia.com/isaac-gym], and run: <pre> tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </pre> 796825471afc45fb7051b90134766358f8e11918 1198 1197 2024-05-22T19:02:56Z Vrtnis 21 /*Stompy expertiments setup*/ wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> Install third-party dependencies: Manually download IsaacGym_Preview_4_Package.tar.gz from [https://developer.nvidia.com/isaac-gym], and run: <pre> tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </pre> == Running Stompy experiments == Download our URDF model from here: <pre> wget https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz && tar -xzvf latest_stl_urdf.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </pre> Run training with the following command: <pre> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </pre> or for full body: <pre> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </pre> Run evaluation with the following command: <pre> python sim/humanoid_gym/play.py --task legs_ppo --sim_device cpu </pre> See this doc for more beginner tips. 23d742d9f7b2fce00d9438128e9d9776440617c2 1199 1198 2024-05-22T19:03:45Z Vrtnis 21 wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> Install third-party dependencies: Manually download IsaacGym_Preview_4_Package.tar.gz from [https://developer.nvidia.com/isaac-gym], and run: <pre> tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </pre> == Running Stompy experiments == Download our URDF model from here: <pre> wget https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz && tar -xzvf latest_stl_urdf.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </pre> Run training with the following command: <pre> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </pre> or for full body: <pre> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </pre> Run evaluation with the following command: <pre> python sim/humanoid_gym/play.py --task legs_ppo --sim_device cpu </pre> See this doc for more beginner tips. == Errors == After cloning Isaac Gym, sometimes the bindings mysteriously disappear. To fix this, update the submodule: <pre> git submodule update --init --recursive </pre> If you observe errors with libpython3.8.so.1.0, you can try the following: <pre> export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH </pre> If you still see segmentation faults, you can try the following: <pre> sudo apt-get install vulkan1 </pre> 77f03eaa2f7774829f00c8e9c7426e3c973aca17 1200 1199 2024-05-22T19:04:02Z Vrtnis 21 /* Errors */ wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> Install third-party dependencies: Manually download IsaacGym_Preview_4_Package.tar.gz from [https://developer.nvidia.com/isaac-gym], and run: <pre> tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </pre> == Running Stompy experiments == Download our URDF model from here: <pre> wget https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz && tar -xzvf latest_stl_urdf.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </pre> Run training with the following command: <pre> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </pre> or for full body: <pre> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </pre> Run evaluation with the following command: <pre> python sim/humanoid_gym/play.py --task legs_ppo --sim_device cpu </pre> See this doc for more beginner tips. == Handling Errors == After cloning Isaac Gym, sometimes the bindings mysteriously disappear. To fix this, update the submodule: <pre> git submodule update --init --recursive </pre> If you observe errors with libpython3.8.so.1.0, you can try the following: <pre> export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH </pre> If you still see segmentation faults, you can try the following: <pre> sudo apt-get install vulkan1 </pre> f95c55665d1eacb8b1e91f924a7b240e58ca2dee 1201 1200 2024-05-22T19:08:27Z Vrtnis 21 wikitext text/x-wiki A library for simulating Stompy in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> Install third-party dependencies: Manually download IsaacGym_Preview_4_Package.tar.gz from [https://developer.nvidia.com/isaac-gym], and run: <pre> tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </pre> == Running Stompy experiments == Download our URDF model from here: <pre> wget https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz && tar -xzvf latest_stl_urdf.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </pre> Run training with the following command: <pre> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </pre> or for full body: <pre> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </pre> Run evaluation with the following command: <pre> python sim/humanoid_gym/play.py --task legs_ppo --sim_device cpu </pre> See this doc for more beginner tips. == Handling Errors == After cloning Isaac Gym, sometimes the bindings mysteriously disappear. To fix this, update the submodule: <pre> git submodule update --init --recursive </pre> If you observe errors with libpython3.8.so.1.0, you can try the following: <pre> export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH </pre> If you still see segmentation faults, you can try the following: <pre> sudo apt-get install vulkan1 </pre> Also, see [[Humanoid Gym]] and [[K-Scale Humanoid Gym]] 560077b140756f9894a60825c146ca9783f5673f 1202 1201 2024-05-22T20:16:23Z Vrtnis 21 /*add wiki-links*/ wikitext text/x-wiki A library for simulating [[Stompy]] in [[Isaac Gym]]. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> Install third-party dependencies: Manually download IsaacGym_Preview_4_Package.tar.gz from [https://developer.nvidia.com/isaac-gym], and run: <pre> tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </pre> == Running Stompy experiments == Download our URDF model from here: <pre> wget https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz && tar -xzvf latest_stl_urdf.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </pre> Run training with the following command: <pre> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </pre> or for full body: <pre> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </pre> Run evaluation with the following command: <pre> python sim/humanoid_gym/play.py --task legs_ppo --sim_device cpu </pre> See this doc for more beginner tips. == Handling Errors == After cloning Isaac Gym, sometimes the bindings mysteriously disappear. To fix this, update the submodule: <pre> git submodule update --init --recursive </pre> If you observe errors with libpython3.8.so.1.0, you can try the following: <pre> export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH </pre> If you still see segmentation faults, you can try the following: <pre> sudo apt-get install vulkan1 </pre> Also, see [[Humanoid Gym]] and [[K-Scale Humanoid Gym]] 0048eb87fd1ae938652db8e91eab7ee73b39f2b1 1203 1202 2024-05-22T20:18:41Z Vrtnis 21 wikitext text/x-wiki A library for simulating [[Stompy]] in Isaac Gym. This library is built on top of the Isaac Gym library and Humanoid-gym and provides a simple interface for running experiments with Stompy. For a start, we have defined two tasks: getting up and walking. The library is available at [https://github.com/kscalelabs/sim https://github.com/kscalelabs/sim] We will be adding more tasks and simulator environments in upcoming weeks. The walking task works reliably with the upper body being fixed. The getting up task is still an open challenge! == Getting Started == This repository requires Python 3.8 due to compatibility issues with underlying libraries. We hope to support more recent Python versions in the future. Clone this repository: <pre> git clone https://github.com/kscalelabs/sim.git cd sim </pre> Create a new conda environment and install the package: <pre> conda create --name kscale-sim-library python=3.8.19 conda activate kscale-sim-library make install-dev </pre> Install third-party dependencies: Manually download IsaacGym_Preview_4_Package.tar.gz from [https://developer.nvidia.com/isaac-gym], and run: <pre> tar -xvf IsaacGym_Preview_4_Package.tar.gz conda env config vars set ISAACGYM_PATH=`pwd`/isaacgym conda deactivate conda activate kscale-sim-library make install-third-party-external </pre> == Running Stompy experiments == Download our URDF model from here: <pre> wget https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz && tar -xzvf latest_stl_urdf.tar.gz python sim/scripts/create_fixed_torso.py export MODEL_DIR=stompy </pre> Run training with the following command: <pre> python sim/humanoid_gym/train.py --task=legs_ppo --num_envs=4096 --headless </pre> or for full body: <pre> python sim/humanoid_gym/train.py --task=stompy_ppo --num_envs=4096 --headless </pre> Run evaluation with the following command: <pre> python sim/humanoid_gym/play.py --task legs_ppo --sim_device cpu </pre> See this doc for more beginner tips. == Handling Errors == After cloning Isaac Gym, sometimes the bindings mysteriously disappear. To fix this, update the submodule: <pre> git submodule update --init --recursive </pre> If you observe errors with libpython3.8.so.1.0, you can try the following: <pre> export LD_LIBRARY_PATH=PATH_TO_YOUR_ENV/lib:$LD_LIBRARY_PATH </pre> If you still see segmentation faults, you can try the following: <pre> sudo apt-get install vulkan1 </pre> Also, see [[Humanoid Gym]] and [[K-Scale Humanoid Gym]] 0c72f0f1fb0bff58e5d7116b9e44f633169f211c 0 278 1204 2024-05-22T20:31:29Z Vrtnis 21 Created page with "== Bi-Manual Remote Robotic Teleoperation == A minimal implementation of a bi-manual remote robotic teleoperation system using VR hand tracking and camera streaming." wikitext text/x-wiki == Bi-Manual Remote Robotic Teleoperation == A minimal implementation of a bi-manual remote robotic teleoperation system using VR hand tracking and camera streaming. fa92490a7759537255b0912659a41e74c0bcd81e 1205 1204 2024-05-22T20:31:40Z Vrtnis 21 /* Bi-Manual Remote Robotic Teleoperation */ wikitext text/x-wiki A minimal implementation of a bi-manual remote robotic teleoperation system using VR hand tracking and camera streaming. c039c8900da4bba16386935311c39f15d295f7d1 1206 1205 2024-05-22T20:33:33Z Vrtnis 21 wikitext text/x-wiki A minimal implementation of a bi-manual remote robotic teleoperation system using VR hand tracking and camera streaming. {| class="wikitable" ! Feature ! Status |- | VR and browser visualization | ✔ Completed |- | Bi-manual hand gesture control | ✔ Completed |- | Camera streaming (mono + stereo) | ✔ Completed |- | Inverse kinematics | ✔ Completed |- | Meta Quest Pro HMD + NVIDIA® Jetson AGX Orin™ Developer Kit | ✔ Completed |- | .urdf robot model | ✔ Completed |- | 3dof end effector control | ✔ Completed |- | Debug 6dof end effector control | ⬜ Pending |- | Resets to various default poses | ⬜ Pending |- | Tested on real world robot | ⬜ Pending |- | Record & playback trajectories | ⬜ Pending |} 33fb743762f4816485d9e7d45f4939f9855c8e0d 1207 1206 2024-05-22T20:34:58Z Vrtnis 21 wikitext text/x-wiki A minimal implementation of a bi-manual remote robotic teleoperation system using VR hand tracking and camera streaming. {| class="wikitable" ! Feature ! Status |- | VR and browser visualization | ✔ Completed |- | Bi-manual hand gesture control | ✔ Completed |- | Camera streaming (mono + stereo) | ✔ Completed |- | Inverse kinematics | ✔ Completed |- | Meta Quest Pro HMD + NVIDIA® Jetson AGX Orin™ Developer Kit | ✔ Completed |- | .urdf robot model | ✔ Completed |- | 3dof end effector control | ✔ Completed |- | Debug 6dof end effector control | ⬜ Pending |- | Resets to various default poses | ⬜ Pending |- | Tested on real world robot | ⬜ Pending |- | Record & playback trajectories | ⬜ Pending |} === Setup === <pre> git clone https://github.com/kscalelabs/teleop.git && cd teleop conda create -y -n teleop python=3.8 && conda activate teleop pip install -r requirements.txt </pre> === Usage === * Start the server on the robot computer: <pre>python demo_hands_stereo_ik3dof.py</pre> * Start ngrok on the robot computer: <pre>ngrok http 8012</pre> * Open the browser app on the HMD and go to the ngrok URL. === Dependencies === * Vuer is used for visualization * PyBullet is used for inverse kinematics * ngrok is used for networking === Citation === <pre> @misc{teleop-2024, title={Bi-Manual Remote Robotic Teleoperation}, author={Hugo Ponte}, year={2024}, url={https://github.com/kscalelabs/teleop} } </pre> 9fcaf48a5af437ebfb8418f489fe2a916fabeea1 0 278 1208 1207 2024-05-22T20:40:58Z Vrtnis 21 Vrtnis moved page [[K-Scale teleop]] to [[K-Scale Teleop]]: Correct title capitalization wikitext text/x-wiki A minimal implementation of a bi-manual remote robotic teleoperation system using VR hand tracking and camera streaming. {| class="wikitable" ! Feature ! Status |- | VR and browser visualization | ✔ Completed |- | Bi-manual hand gesture control | ✔ Completed |- | Camera streaming (mono + stereo) | ✔ Completed |- | Inverse kinematics | ✔ Completed |- | Meta Quest Pro HMD + NVIDIA® Jetson AGX Orin™ Developer Kit | ✔ Completed |- | .urdf robot model | ✔ Completed |- | 3dof end effector control | ✔ Completed |- | Debug 6dof end effector control | ⬜ Pending |- | Resets to various default poses | ⬜ Pending |- | Tested on real world robot | ⬜ Pending |- | Record & playback trajectories | ⬜ Pending |} === Setup === <pre> git clone https://github.com/kscalelabs/teleop.git && cd teleop conda create -y -n teleop python=3.8 && conda activate teleop pip install -r requirements.txt </pre> === Usage === * Start the server on the robot computer: <pre>python demo_hands_stereo_ik3dof.py</pre> * Start ngrok on the robot computer: <pre>ngrok http 8012</pre> * Open the browser app on the HMD and go to the ngrok URL. === Dependencies === * Vuer is used for visualization * PyBullet is used for inverse kinematics * ngrok is used for networking === Citation === <pre> @misc{teleop-2024, title={Bi-Manual Remote Robotic Teleoperation}, author={Hugo Ponte}, year={2024}, url={https://github.com/kscalelabs/teleop} } </pre> 9fcaf48a5af437ebfb8418f489fe2a916fabeea1 0 279 1209 2024-05-22T20:40:58Z Vrtnis 21 Vrtnis moved page [[K-Scale teleop]] to [[K-Scale Teleop]]: Correct title capitalization wikitext text/x-wiki #REDIRECT [[K-Scale Teleop]] 214fcb97c3e517152059dec71c45f1a5f75d7a80 0 280 1210 2024-05-22T20:43:18Z Vrtnis 21 Created page with "== Setup - Linux == '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.1..." wikitext text/x-wiki == Setup - Linux == '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip pip install -e . </pre> '''Run tests''' <pre> pip install pytest pytest tests/test_env.py </pre> 8c36c9a3b242b7fc630e2bbe089e76cc59dab77f 1211 1210 2024-05-22T20:44:44Z Vrtnis 21 wikitext text/x-wiki == Setup - Linux == '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip pip install -e . </pre> '''Run tests''' <pre> pip install pytest pytest tests/test_env.py </pre> == Setup - Jetson Orin AGX == '''No conda on ARM64, install on bare metal''' <pre> sudo apt-get install libhdf5-dev git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip pip install -e . </pre> 7f044f04386d1d5efe84d47c4ec3b9a561ebbec9 1212 1211 2024-05-22T20:47:43Z Vrtnis 21 wikitext text/x-wiki == Setup - Linux == '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip pip install -e . </pre> '''Run tests''' <pre> pip install pytest pytest tests/test_env.py </pre> == Setup - Jetson Orin AGX == '''No conda on ARM64, install on bare metal''' <pre> sudo apt-get install libhdf5-dev git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip pip install -e . </pre> == Usage - Basic == '''Visualize the MuJoCo scene''' <pre> python gym_kmanip/examples/1_view_env.py </pre> '''Record a video of the MuJoCo scene''' <pre> python gym_kmanip/examples/2_record_video.py </pre> == Usage - Recording Data == '''K-Scale HuggingFace Datasets''' '''Data is recorded via teleop, this requires additional dependencies''' <pre> pip install opencv-python==4.9.0.80 pip install vuer==0.0.30 pip install rerun-sdk==0.16.0 </pre> '''Start the server on the robot computer''' <pre> python gym_kmanip/examples/4_record_data_teleop.py </pre> '''Start ngrok on the robot computer''' <pre> ngrok http 8012 </pre> '''Open the browser app on the VR headset and go to the ngrok URL''' 715b688e127b9bd990789b006a78851fa4f9a925 1213 1212 2024-05-22T20:48:12Z Vrtnis 21 wikitext text/x-wiki == Setup - Linux == '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip pip install -e . </pre> '''Run tests''' <pre> pip install pytest pytest tests/test_env.py </pre> == Setup - Jetson Orin AGX == '''No conda on ARM64, install on bare metal''' <pre> sudo apt-get install libhdf5-dev git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip pip install -e . </pre> == Usage - Basic == '''Visualize the MuJoCo scene''' <pre> python gym_kmanip/examples/1_view_env.py </pre> '''Record a video of the MuJoCo scene''' <pre> python gym_kmanip/examples/2_record_video.py </pre> == Usage - Recording Data == '''K-Scale HuggingFace Datasets''' '''Data is recorded via teleop, this requires additional dependencies''' <pre> pip install opencv-python==4.9.0.80 pip install vuer==0.0.30 pip install rerun-sdk==0.16.0 </pre> '''Start the server on the robot computer''' <pre> python gym_kmanip/examples/4_record_data_teleop.py </pre> '''Start ngrok on the robot computer''' <pre> ngrok http 8012 </pre> '''Open the browser app on the VR headset and go to the ngrok URL''' == Usage - Visualizing Data == '''Data is visualized using rerun''' <pre> rerun gym_kmanip/data/test.rrd </pre> == Usage - MuJoCo Sim Visualizer == '''MuJoCo provides a nice visualizer where you can directly control the robot''' '''Download standalone MuJoCo''' <pre> tar -xzf ~/Downloads/mujoco-3.1.5-linux-x86_64.tar.gz -C /path/to/mujoco-3.1.5 </pre> '''Run the simulator''' <pre> /path/to/mujoco-3.1.5/bin/simulate gym_kmanip/assets/_env_solo_arm.xml </pre> 792c28c85f7b9acafa137b7f38eefb915a5e5164 1214 1213 2024-05-22T20:49:57Z Vrtnis 21 wikitext text/x-wiki == Setup - Linux == '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip pip install -e . </pre> '''Run tests''' <pre> pip install pytest pytest tests/test_env.py </pre> == Setup - Jetson Orin AGX == '''No conda on ARM64, install on bare metal''' <pre> sudo apt-get install libhdf5-dev git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip pip install -e . </pre> == Usage - Basic == '''Visualize the MuJoCo scene''' <pre> python gym_kmanip/examples/1_view_env.py </pre> '''Record a video of the MuJoCo scene''' <pre> python gym_kmanip/examples/2_record_video.py </pre> == Usage - Recording Data == '''K-Scale HuggingFace Datasets''' '''Data is recorded via teleop, this requires additional dependencies''' <pre> pip install opencv-python==4.9.0.80 pip install vuer==0.0.30 pip install rerun-sdk==0.16.0 </pre> '''Start the server on the robot computer''' <pre> python gym_kmanip/examples/4_record_data_teleop.py </pre> '''Start ngrok on the robot computer''' <pre> ngrok http 8012 </pre> '''Open the browser app on the VR headset and go to the ngrok URL''' == Usage - Visualizing Data == '''Data is visualized using rerun''' <pre> rerun gym_kmanip/data/test.rrd </pre> == Usage - MuJoCo Sim Visualizer == '''MuJoCo provides a nice visualizer where you can directly control the robot''' '''Download standalone MuJoCo''' <pre> tar -xzf ~/Downloads/mujoco-3.1.5-linux-x86_64.tar.gz -C /path/to/mujoco-3.1.5 </pre> '''Run the simulator''' <pre> /path/to/mujoco-3.1.5/bin/simulate gym_kmanip/assets/_env_solo_arm.xml </pre> <pre> Citation: @misc{teleop-2024, title={gym-kmanip}, author={Hugo Ponte}, year={2024}, url={https://github.com/kscalelabs/gym-kmanip} } </pre> 097728d67ba52ff1f328990947267b7dbcb4bf17 1215 1214 2024-05-22T20:50:31Z Vrtnis 21 wikitext text/x-wiki == Setup - Linux == '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip pip install -e . </pre> '''Run tests''' <pre> pip install pytest pytest tests/test_env.py </pre> == Setup - Jetson Orin AGX == '''No conda on ARM64, install on bare metal''' <pre> sudo apt-get install libhdf5-dev git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip pip install -e . </pre> == Usage - Basic == '''Visualize the MuJoCo scene''' <pre> python gym_kmanip/examples/1_view_env.py </pre> '''Record a video of the MuJoCo scene''' <pre> python gym_kmanip/examples/2_record_video.py </pre> == Usage - Recording Data == '''K-Scale HuggingFace Datasets''' '''Data is recorded via teleop, this requires additional dependencies''' <pre> pip install opencv-python==4.9.0.80 pip install vuer==0.0.30 pip install rerun-sdk==0.16.0 </pre> '''Start the server on the robot computer''' <pre> python gym_kmanip/examples/4_record_data_teleop.py </pre> '''Start ngrok on the robot computer''' <pre> ngrok http 8012 </pre> '''Open the browser app on the VR headset and go to the ngrok URL''' == Usage - Visualizing Data == '''Data is visualized using rerun''' <pre> rerun gym_kmanip/data/test.rrd </pre> == Usage - MuJoCo Sim Visualizer == '''MuJoCo provides a nice visualizer where you can directly control the robot''' '''Download standalone MuJoCo''' <pre> tar -xzf ~/Downloads/mujoco-3.1.5-linux-x86_64.tar.gz -C /path/to/mujoco-3.1.5 </pre> '''Run the simulator''' <pre> /path/to/mujoco-3.1.5/bin/simulate gym_kmanip/assets/_env_solo_arm.xml </pre> == Citation == <pre> @misc{teleop-2024, title={gym-kmanip}, author={Hugo Ponte}, year={2024}, url={https://github.com/kscalelabs/gym-kmanip} } </pre> 82af190b19d1fdc37439d0087f4cbf5ff8be9224 1216 1215 2024-05-22T20:51:32Z Vrtnis 21 wikitext text/x-wiki == Setup - Linux == [https://github.com/kscalelabs/gym-kmanip https://github.com/kscalelabs/gym-kmanip] '''Clone and install dependencies''' <pre> git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip pip install -e . </pre> '''Run tests''' <pre> pip install pytest pytest tests/test_env.py </pre> == Setup - Jetson Orin AGX == '''No conda on ARM64, install on bare metal''' <pre> sudo apt-get install libhdf5-dev git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip pip install -e . </pre> == Usage - Basic == '''Visualize the MuJoCo scene''' <pre> python gym_kmanip/examples/1_view_env.py </pre> '''Record a video of the MuJoCo scene''' <pre> python gym_kmanip/examples/2_record_video.py </pre> == Usage - Recording Data == '''K-Scale HuggingFace Datasets''' '''Data is recorded via teleop, this requires additional dependencies''' <pre> pip install opencv-python==4.9.0.80 pip install vuer==0.0.30 pip install rerun-sdk==0.16.0 </pre> '''Start the server on the robot computer''' <pre> python gym_kmanip/examples/4_record_data_teleop.py </pre> '''Start ngrok on the robot computer''' <pre> ngrok http 8012 </pre> '''Open the browser app on the VR headset and go to the ngrok URL''' == Usage - Visualizing Data == '''Data is visualized using rerun''' <pre> rerun gym_kmanip/data/test.rrd </pre> == Usage - MuJoCo Sim Visualizer == '''MuJoCo provides a nice visualizer where you can directly control the robot''' '''Download standalone MuJoCo''' <pre> tar -xzf ~/Downloads/mujoco-3.1.5-linux-x86_64.tar.gz -C /path/to/mujoco-3.1.5 </pre> '''Run the simulator''' <pre> /path/to/mujoco-3.1.5/bin/simulate gym_kmanip/assets/_env_solo_arm.xml </pre> == Citation == <pre> @misc{teleop-2024, title={gym-kmanip}, author={Hugo Ponte}, year={2024}, url={https://github.com/kscalelabs/gym-kmanip} } </pre> 21d1535a2c2a1e74c746adaf0b5e1e967b64b5d6 0 257 1217 1160 2024-05-23T17:01:19Z Kewang 11 wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == First git clone <code> https://github.com/zalo/mujoco_wasm </code> Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Running in Browser == Run this in your mujoco folder to start a server. <code> python -m http.server 8000 </code> Then navigate to: <code> http://localhost:8000/index.html </code> [[File:Wasm screenshot13-40-40.png|800px|thumb|none|MuJoCo running in browser]] == Running in Cloud/Cluster and Viewing on Local Machine == Add extra parameter to your ssh command: <code> ssh -L 8000:127.0.0.1:8000 my_name@my_cluster_ip </code> Then you can open it on the browser on your local machine! d7e990e7a8d776a82813813c91664a405ff29177 0 281 1218 2024-05-23T20:51:06Z Vrtnis 21 Created page with "=== Educational === {| class="wikitable" |- ! Name !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || [URDF](https://github.com/Gepetto/example-..." wikitext text/x-wiki === Educational === {| class="wikitable" |- ! Name !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || [URDF](https://github.com/laas/simple_humanoid_description) || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || [URDF](https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description) || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} |} e5373f7aa5ad10d20d71e50567b4782d6b294ed7 1219 1218 2024-05-23T20:51:15Z Vrtnis 21 /* Educational */ wikitext text/x-wiki === Educational === {| class="wikitable" |- ! Name !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || [URDF](https://github.com/laas/simple_humanoid_description) || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || [URDF](https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description) || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || [URDF](https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf), [MJCF](https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro) || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand) || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85), [URDF](https://github.com/a-price/robotiq_arg85_description), [Xacro](https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization) || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || [URDF](https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots) || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || [SDF](https://github.com/RobotLocomotion/models/tree/master/wsg_50_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 1dfb184837d971ddd6edc56beb6acec9eeeb796a 1220 1219 2024-05-23T21:22:32Z Vrtnis 21 wikitext text/x-wiki ** updates in progress ** === Educational === {| class="wikitable" |- ! Name !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || [URDF](https://github.com/laas/simple_humanoid_description) || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || [URDF](https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description) || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || [URDF](https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf), [MJCF](https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro) || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand) || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85), [URDF](https://github.com/a-price/robotiq_arg85_description), [Xacro](https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization) || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || [URDF](https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots) || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || [SDF](https://github.com/RobotLocomotion/models/tree/master/wsg_50_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 544cd048c05f4ff818016ceea21ff5f5ee62c165 1221 1220 2024-05-23T21:22:53Z Vrtnis 21 wikitext text/x-wiki == ** updates in progress ** == === Educational === {| class="wikitable" |- ! Name !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || [URDF](https://github.com/laas/simple_humanoid_description) || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || [URDF](https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description) || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || [URDF](https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf), [MJCF](https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro) || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand) || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85), [URDF](https://github.com/a-price/robotiq_arg85_description), [Xacro](https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization) || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || [URDF](https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots) || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || [SDF](https://github.com/RobotLocomotion/models/tree/master/wsg_50_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} cb120f09e43108c29e5cc916543ddae09558e4df 1222 1221 2024-05-24T00:54:23Z Vrtnis 21 wikitext text/x-wiki == ** updates in progress ** == === Educational === {| class="wikitable" |- ! Name !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [URDF](https://github.com/laas/simple_humanoid_description) || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [URDF](https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description) || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [URDF](https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [URDF](https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf), [MJCF](https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro) || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand) || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [MJCF](https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85), [URDF](https://github.com/a-price/robotiq_arg85_description), [Xacro](https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization) || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [URDF](https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots) || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [SDF](https://github.com/RobotLocomotion/models/tree/master/wsg_50_description) || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} a889d9c3fd779eadddcef184e5976e9bc1aacade 1223 1222 2024-05-24T01:10:35Z Vrtnis 21 wikitext text/x-wiki == ** updates in progress ** == === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} d824eee2d5cf4d60f44998f60ac18b2bc10d7a4d 1224 1223 2024-05-24T01:11:44Z Vrtnis 21 wikitext text/x-wiki == ** updates in progress ** == === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} e2311e97f9a0879892e8ab1522fb197ef6369436 1225 1224 2024-05-24T01:16:20Z Vrtnis 21 wikitext text/x-wiki === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} ac3bf2e3557cb254bab4bfa2c126da22b7f39009 1226 1225 2024-05-24T01:19:41Z Vrtnis 21 /* Add drones */ wikitext text/x-wiki === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Drones === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | X2 || Skydio || MJCF, URDF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/skydio_x2 MJCF], [https://github.com/lvjonok/skydio_x2_description/blob/master/urdf/skydio_x2.urdf URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Crazyflie 2.0 || Bitcraze || URDF, MJCF || [https://github.com/utiasDSL/gym-pybullet-drones/tree/master/gym_pybullet_drones/assets URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/bitcraze_crazyflie_2 MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Crazyflie 1.0 || Bitcraze || Xacro || [https://github.com/whoenig/crazyflie_ros/tree/master/crazyflie_description Xacro] || MIT || ✔️ || ✔️ || ✖️ |} ca443b85a6d22ae008ff91bd2251f95b9490d638 1227 1226 2024-05-24T01:27:03Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Drones === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | X2 || Skydio || MJCF, URDF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/skydio_x2 MJCF], [https://github.com/lvjonok/skydio_x2_description/blob/master/urdf/skydio_x2.urdf URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Crazyflie 2.0 || Bitcraze || URDF, MJCF || [https://github.com/utiasDSL/gym-pybullet-drones/tree/master/gym_pybullet_drones/assets URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/bitcraze_crazyflie_2 MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Crazyflie 1.0 || Bitcraze || Xacro || [https://github.com/whoenig/crazyflie_ros/tree/master/crazyflie_description Xacro] || MIT || ✔️ || ✔️ || ✖️ |} bbc54f69bf28162ca4128e4963b718b0ad812481 1228 1227 2024-05-24T01:27:35Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 98a1eb7ff17908e63052b7267d6d8176529c852f 1229 1228 2024-05-24T01:27:55Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://media.kscale.dev/stompy/latest_stl_urdf.tar.gz URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 6038fa81cefc3547ef07c13804d9a8b329a54e2d 1230 1229 2024-05-24T02:14:28Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 53cd290136aee83c13bdcc99284dc34d1e823c43 0 270 1231 1167 2024-05-24T20:07:08Z 108.211.178.220 0 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy pi that maps states or observations to actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a network of conditional probability, we have * <math> s_1 -> o_1 - (\pi_\theta) -> a_1 </math> (policy) * <math> s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 </math> (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. We often use and tune hyperparameters for reward functions to make model training faster === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} </math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} </math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let <math> T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k) </math>. r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat Policy Gradients - Directly differentiate objective with respect to the optimal theta and then perform gradient descent Value-based: Estimate value function or q-function of optimal policy (policy is often represented implicitly) Actor-Critic: Estimate value function or q-function of current policy, and find a better policy gradient Model-based: Estimate some transition model, and then use it to improve a policy === REINFORCE === - === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here <math>\alpha</math> represents the learning rate, which is how much new information is weighted relative to old information, while <math>\gamma</math> represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. <math> Q_{i+1} (s,a) = (1 - \alpha) Q_i (s,a) + \alpha (r(s,a) + \gamma V_i(s'))</math> Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. d8b4ab53a0b25aa0c9633209de324325f8a35bf2 1232 1231 2024-05-24T20:09:45Z Allen12 15 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy <math>pi</math> that maps states or observations to optimal actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a network of conditional probability, we have * <math> s_1 -> o_1 - (\pi_\theta) -> a_1 </math> (policy) * <math> s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 </math> (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. We often use and tune hyperparameters for reward functions to make model training faster === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} </math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} </math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let <math> T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k) </math>. r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat Policy Gradients - Directly differentiate objective with respect to the optimal theta and then perform gradient descent Value-based: Estimate value function or q-function of optimal policy (policy is often represented implicitly) Actor-Critic: Estimate value function or q-function of current policy, and find a better policy gradient Model-based: Estimate some transition model, and then use it to improve a policy === REINFORCE === - === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here <math>\alpha</math> represents the learning rate, which is how much new information is weighted relative to old information, while <math>\gamma</math> represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. <math> Q_{i+1} (s,a) = (1 - \alpha) Q_i (s,a) + \alpha (r(s,a) + \gamma V_i(s'))</math> Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. c2b41c8f012378f234ab70ff016f29778e0673c6 1250 1232 2024-05-24T23:55:35Z Allen12 15 wikitext text/x-wiki Allen's reinforcement learning notes === Links === * [https://rail.eecs.berkeley.edu/deeprlcourse-fa19/ Berkeley CS285] * [https://www.youtube.com/watch?v=SupFHGbytvA&list=PL_iWQOsE6TfVYGEGiAOMaOzzv41Jfm_Ps Sergey Levine RL Lecture] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy <math>pi</math> that maps states or observations to optimal actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always know what s_t is from o_t, but o_t is inferable from s_t. To think of it as a network of conditional probability, we have * <math> s_1 -> o_1 - (\pi_\theta) -> a_1 </math> (policy) * <math> s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 </math> (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. We often use and tune hyperparameters for reward functions to make model training faster === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} </math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} </math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let <math> T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k) </math>. r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat Policy Gradients - Directly differentiate objective with respect to the optimal theta and then perform gradient descent Value-based: Estimate value function or q-function of optimal policy (policy is often represented implicitly) Actor-Critic: Estimate value function or q-function of current policy, and find a better policy gradient Model-based: Estimate some transition model, and then use it to improve a policy === REINFORCE === - === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here <math>\alpha</math> represents the learning rate, which is how much new information is weighted relative to old information, while <math>\gamma</math> represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. <math> Q_{i+1} (s,a) = (1 - \alpha) Q_i (s,a) + \alpha (r(s,a) + \gamma V_i(s'))</math> Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. 62e08d2e5aabfb712b9bc7966fd45f5c90251eb3 0 282 1233 2024-05-24T20:11:06Z Allen12 15 Created page with "Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf] [[Category:Reinforcement Learning]] === Motivation === Consider a problem wh..." wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf] [[Category:Reinforcement Learning]] === Motivation === Consider a problem where we have to train a robot to pick up some object. A traditional ML algorithm might try to learn some function f(x) = y, where given some position x observed via the camera we output some behavior y. The trouble is that in the real world, the correct grab location is some function of the object and the physical environment, which is hard to intuitively ascertain by observation. The motivation behind reinforcement learning is to repeatedly take observations, then sample the effects of actions on those observations (reward and new observation/state). Ultimately, we hope to create a policy <math>pi</math> that maps states or observations to optimal actions. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === A state is a complete representation of the physical world while the observation is some subset or representation of s. They are not necessarily the same in that we can't always infer s_t from o_t, but o_t is inferable from s_t. To think of it as a network of conditional probability, we have * <math> s_1 -> o_1 - (\pi_\theta) -> a_1 </math> (policy) * <math> s_1, a_1 - (p(s_{t+1} | s_t, a_t) -> s_2 </math> (dynamics) Note that theta represents the parameters of the policy (for example, the parameters of a neural network). Assumption: Markov Property - Future states are independent of past states given present states. This is the fundamental difference between states and observations. === Problem Representation === States and actions are typically continuous - thus, we often want to model our output policy as a density function, which tells us the distribution of probabilities of actions at some given state. The reward is a function of the state and action r(s, a) -> int, which tells us what states and actions are better. We often use and tune hyperparameters for reward functions to make model training faster === Markov Chain & Decision Process=== Markov Chain: <math> M = {S, T} </math>, where S - state space, T- transition operator. The state space is the set of all states, and can be discrete or continuous. The transition probabilities is represented in a matrix, where the i,j'th entry is the probability of going into state i at state j, and we can express the next time step by multiplying the current time step with the transition operator. Markov Decision Process: <math> M = {S, A, T, r} </math>, where A - action space. T is now a tensor, containing the current state, current action, and next state. We let <math> T_{i, j, k} = p(s_t + 1 = i | s_t = j, a_t = k) </math>. r is the reward function. === Reinforcement Learning Algorithms - High-level === # Generate Samples (run policy) # Fit a model/estimate something about how well policy is performing # Improve policy # Repeat Policy Gradients - Directly differentiate objective with respect to the optimal theta and then perform gradient descent Value-based: Estimate value function or q-function of optimal policy (policy is often represented implicitly) Actor-Critic: Estimate value function or q-function of current policy, and find a better policy gradient Model-based: Estimate some transition model, and then use it to improve a policy === REINFORCE === - === Temporal Difference Learning === Temporal Difference (TD) is a model for estimating the utility of states given some state-action-outcome information. Suppose we have some initial value <math>V_0(s) </math>, and we get some information <math> (s, a, s', r(s, a) </math>. We can then use the update equation <math>V_{t+1}(s) = (1- \alpha)V_{t}(s)+\alpha(R(s, a, s') + \gamma V_i(s')) </math>. Here <math>\alpha</math> represents the learning rate, which is how much new information is weighted relative to old information, while <math>\gamma</math> represents the discount factor, which can be thought of how much getting a reward in the future factors into our current reward. === Q Learning === Q Learning gives us a way to extract the optimal policy after learning. Instead of keeping track of the values of individual states, we keep track of Q values for state-action pairs, representing the utility of taking action a at state s. How do we use this Q value? Two main ideas. Idea 1: Policy iteration - if we have a policy <math> \pi </math> and we know <math> Q^pi (s, a) </math>, we can improve the policy, by deterministically setting the action at each state be the argmax of all possible actions at the state. <math> Q_{i+1} (s,a) = (1 - \alpha) Q_i (s,a) + \alpha (r(s,a) + \gamma V_i(s'))</math> Idea 2: Gradient update - If <math> Q^pi(s, a) > V^pi(s) </math>, then a is better than average. We will then modify the policy to increase the probability of a. 86b2547c7288b924baa6045775913215963663a2 1234 1233 2024-05-24T20:11:25Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf] [[Category:Reinforcement Learning]] === Motivation === === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === 4427f0ba1681a908e5a390862d6ba2379b31ae3e 1235 1234 2024-05-24T20:24:12Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf /RLbook2020] [[Category:Reinforcement Learning]] === Motivation === === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === fc04d2a33d3a64b26cbf95d98c19a19b2d10ddc9 1236 1235 2024-05-24T20:24:22Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === fb450ca3cb73af7221f1e6713c4b667c94dc8416 1237 1236 2024-05-24T21:41:42Z Allen12 15 /* Motivation */ wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math>\pi^*<\math> which we encode in a neural network with parameters <math>\theta^*<\math>. These optimal parameters are defined as <math>\theta^* = \text<argmax>_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] <\math> === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === cf8c15f7202d4ad8a6ee6c349f0fc683ad61c030 1238 1237 2024-05-24T21:42:01Z Allen12 15 /* Motivation */ wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math>\pi^*<\math> which we encode in a neural network with parameters <math>\theta^*</math>. These optimal parameters are defined as <math>\theta^* = \text<argmax>_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math> === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === 64d04a9b522c8c98f7bd207b5bddd45b5e36b97b 1239 1238 2024-05-24T21:42:21Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. These optimal parameters are defined as <math>\theta^* = \text<argmax>_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math> === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === dc482ab67855780225cdd65199e9e32a0f20f0c2 1240 1239 2024-05-24T21:43:50Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math> === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === afa2a2a0629c983a8716b35019d6b221dc436e02 1241 1240 2024-05-24T21:46:04Z Allen12 15 /* Motivation */ wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory determined by the policy is the highest over all policies. === Learning === Learning involves the agent taking actions and the environment returning a new state and reward. * Input: <math>s_t</math>: States at each time step * Output: <math>a_t</math>: Actions at each time step * Data: <math>(s_1, a_1, r_1, ... , s_T, a_T, r_T)</math> * Learn <math>\pi_\theta : s_t -> a_t </math> to maximize <math> \sum_t r_t </math> === State vs. Observation === 5d5ff0def524373ab8faea25bc516cf0fc62d168 1251 1241 2024-05-24T23:58:17Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === # Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions. Remember a policy is a mapping from observations to outputs. If the space is continuous, it may make more sense to make output be one mean and one standard deviation for each component of the action. # Repeat: === State vs. Observation === b06fbf05ed8ed2344a3ef915eddf013135276bb3 1252 1251 2024-05-25T00:03:34Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === # Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions. Remember a policy is a mapping from observations to outputs. If the space is continuous, it may make more sense to make output be one mean and one standard deviation for each component of the action. ‎<syntaxhighlight lang="python" line> # For # of episodes: ## While not terminated: ### Get observation from environment ### Use policy network to map observation to action distribution ### Randomly sample one action from action distribution ### Compute logarithmic probability of that action occurring ### Step environment using action and store reward ## Calculate loss over entire trajectory as function of probabilities and rewards </syntaxhighlight> === Loss Function === c9e07c8cf509e9f20c1e46916b0ce522df3d288d 1253 1252 2024-05-25T00:05:30Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" line> Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For # of episodes: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards </syntaxhighlight> === Loss Function === 94090e441140e894774e82de1ca44232ec590fad 1254 1253 2024-05-25T00:05:44Z Allen12 15 /* Overview */ wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" line> Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For \# of episodes: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards </syntaxhighlight> === Loss Function === 2962a5d181a846e810724c04e71deb7522bb34d4 0 283 1242 2024-05-24T23:39:45Z Vedant 24 Created page with "=== CAN Daddy === == Design Requirements ==" wikitext text/x-wiki === CAN Daddy === == Design Requirements == 52a6aaea9cae6a897b368d0991480b197c9d2948 1243 1242 2024-05-24T23:39:53Z Vedant 24 /* CAN Daddy */ wikitext text/x-wiki == Design Requirements == bd8492fb86d2336f8b4e0db4807fd4489c171f4d 1244 1243 2024-05-24T23:40:05Z Vedant 24 /* Design Requirements */ wikitext text/x-wiki == Design Requirements == *Hello there* 868027d6f003fdca4af4d8cf08c43242a634630a 1245 1244 2024-05-24T23:40:13Z Vedant 24 /* Design Requirements */ wikitext text/x-wiki == Design Requirements == *Hello there asdf 15a533bf51098cfa5d682d55b0869604cb10d26c 1246 1245 2024-05-24T23:40:22Z Vedant 24 /* Design Requirements */ wikitext text/x-wiki == Design Requirements == *Hello there d7a5bff3a94654e6334a2803e10a759200b5e791 1247 1246 2024-05-24T23:40:55Z Vedant 24 /* Design Requirements */ wikitext text/x-wiki == Design Requirements == * Means bullet point = * n means a header of some kind. Need to wrap it. 73f5f5fa681ea7cfb2a2c8675a1345ebf4e97381 0 16 1248 1040 2024-05-24T23:51:44Z Ben 2 /* Andromeda Cluster */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. Don't do anything computationally expensive on the main node or you will crash it for everyone. Instead, when you need to run some experiments, reserve a GPU (see below). ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. ==== Useful Commands ==== Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] a96941b1701791c89cc07584a7c12532e8ff6816 1249 1248 2024-05-24T23:55:00Z Ben 2 /* Andromeda Cluster */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. Don't do anything computationally expensive on the main node or you will crash it for everyone. Instead, when you need to run some experiments, reserve a GPU (see below). ==== SLURM Commands ==== Show all currently running jobs: <syntaxhighlight lang="bash"> squeue </syntaxhighlight> Show your own running jobs: <syntaxhighlight lang="bash"> squeue --me </syntaxhighlight> Show the available partitions on the cluster: <syntaxhighlight lang="bash"> sinfo </syntaxhighlight> You'll see something like this: <syntaxhighlight lang="bash"> $ sinfo PARTITION AVAIL TIMELIMIT NODES STATE NODELIST compute* up infinite 8 idle compute-permanent-node-[68,285,493,580,625-626,749,801] </syntaxhighlight> This means: * There is one compute node type, called <code>compute</code> * There are 8 nodes of that type, all currently in <code>idle</code> state * The node names are things like <code>compute-permanent-node-68</code> ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. ==== Useful Commands ==== Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] ba245573d96b2a29375532ad5e339aedfcc58286 1255 1249 2024-05-25T00:06:27Z Ben 2 /* SLURM Commands */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. Don't do anything computationally expensive on the main node or you will crash it for everyone. Instead, when you need to run some experiments, reserve a GPU (see below). ==== SLURM Commands ==== Show all currently running jobs: <syntaxhighlight lang="bash"> squeue </syntaxhighlight> Show your own running jobs: <syntaxhighlight lang="bash"> squeue --me </syntaxhighlight> Show the available partitions on the cluster: <syntaxhighlight lang="bash"> sinfo </syntaxhighlight> You'll see something like this: <syntaxhighlight lang="bash"> $ sinfo PARTITION AVAIL TIMELIMIT NODES STATE NODELIST compute* up infinite 8 idle compute-permanent-node-[68,285,493,580,625-626,749,801] </syntaxhighlight> This means: * There is one compute node type, called <code>compute</code> * There are 8 nodes of that type, all currently in <code>idle</code> state * The node names are things like <code>compute-permanent-node-68</code> To launch a job, use [https://slurm.schedmd.com/srun.html srun] or [https://slurm.schedmd.com/sbatch.html sbatch]. * '''srun''' runs a command directly with the requested resources * '''sbatch''' queues the job to run when resources become available For example, suppose I have the following Shell script: <syntaxhighlight lang="bash"> #!/bin/bash echo "Hello, world!" nvidia-smi </syntaxhighlight> I can use <code>srun</code> to run this script with the following result: <syntaxhighlight lang="bash"> $ srun --gpus 8 ./test.sh Hello, world! Sat May 25 00:02:23 2024 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 550.54.15 Driver Version: 550.54.15 CUDA Version: 12.4 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| ... truncated </syntaxhiglight> Alternatively, I can queue the job using <code>sbatch</code>, which gives me the following result: <syntaxhighlight lang="bash"> $ sbatch test.sh Submitted batch job 461 </syntaxhighlight> After launching the job, we can see it running using our original <code>squeue</code> command: <syntaxhighlight lang="bash"> $ squeue --me JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 461 compute test.sh ben R 0:37 1 compute-permanent-node-285 </syntaxhighlight> ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. ==== Useful Commands ==== Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] b121c42578368944e5293d5c0defca46b1fa0b2d 1256 1255 2024-05-25T00:07:19Z Ben 2 /* SLURM Commands */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. Don't do anything computationally expensive on the main node or you will crash it for everyone. Instead, when you need to run some experiments, reserve a GPU (see below). ==== SLURM Commands ==== Show all currently running jobs: <syntaxhighlight lang="bash"> squeue </syntaxhighlight> Show your own running jobs: <syntaxhighlight lang="bash"> squeue --me </syntaxhighlight> Show the available partitions on the cluster: <syntaxhighlight lang="bash"> sinfo </syntaxhighlight> You'll see something like this: <syntaxhighlight lang="bash"> $ sinfo PARTITION AVAIL TIMELIMIT NODES STATE NODELIST compute* up infinite 8 idle compute-permanent-node-[68,285,493,580,625-626,749,801] </syntaxhighlight> This means: * There is one compute node type, called <code>compute</code> * There are 8 nodes of that type, all currently in <code>idle</code> state * The node names are things like <code>compute-permanent-node-68</code> To launch a job, use [https://slurm.schedmd.com/srun.html srun] or [https://slurm.schedmd.com/sbatch.html sbatch]. * '''srun''' runs a command directly with the requested resources * '''sbatch''' queues the job to run when resources become available For example, suppose I have the following Shell script: <syntaxhighlight lang="bash"> #!/bin/bash echo "Hello, world!" nvidia-smi </syntaxhighlight> I can use <code>srun</code> to run this script with the following result: <syntaxhighlight lang="bash"> $ srun --gpus 8 ./test.sh Hello, world! Sat May 25 00:02:23 2024 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 550.54.15 Driver Version: 550.54.15 CUDA Version: 12.4 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| ... truncated </syntaxhighlight> Alternatively, I can queue the job using <code>sbatch</code>, which gives me the following result: <syntaxhighlight lang="bash"> $ sbatch test.sh Submitted batch job 461 </syntaxhighlight> After launching the job, we can see it running using our original <code>squeue</code> command: <syntaxhighlight lang="bash"> $ squeue --me JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 461 compute test.sh ben R 0:37 1 compute-permanent-node-285 </syntaxhighlight> ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. ==== Useful Commands ==== Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] 28f6d95765ab8c70f7b5ba365d84a4f95646c125 1257 1256 2024-05-25T00:10:45Z Ben 2 /* SLURM Commands */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. Don't do anything computationally expensive on the main node or you will crash it for everyone. Instead, when you need to run some experiments, reserve a GPU (see below). ==== SLURM Commands ==== Show all currently running jobs: <syntaxhighlight lang="bash"> squeue </syntaxhighlight> Show your own running jobs: <syntaxhighlight lang="bash"> squeue --me </syntaxhighlight> Show the available partitions on the cluster: <syntaxhighlight lang="bash"> sinfo </syntaxhighlight> You'll see something like this: <syntaxhighlight lang="bash"> $ sinfo PARTITION AVAIL TIMELIMIT NODES STATE NODELIST compute* up infinite 8 idle compute-permanent-node-[68,285,493,580,625-626,749,801] </syntaxhighlight> This means: * There is one compute node type, called <code>compute</code> * There are 8 nodes of that type, all currently in <code>idle</code> state * The node names are things like <code>compute-permanent-node-68</code> To launch a job, use [https://slurm.schedmd.com/srun.html srun] or [https://slurm.schedmd.com/sbatch.html sbatch]. * '''srun''' runs a command directly with the requested resources * '''sbatch''' queues the job to run when resources become available For example, suppose I have the following Shell script: <syntaxhighlight lang="bash"> #!/bin/bash echo "Hello, world!" nvidia-smi </syntaxhighlight> I can use <code>srun</code> to run this script with the following result: <syntaxhighlight lang="bash"> $ srun --gpus 8 ./test.sh Hello, world! Sat May 25 00:02:23 2024 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 550.54.15 Driver Version: 550.54.15 CUDA Version: 12.4 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| ... truncated </syntaxhighlight> Alternatively, I can queue the job using <code>sbatch</code>, which gives me the following result: <syntaxhighlight lang="bash"> $ sbatch test.sh Submitted batch job 461 </syntaxhighlight> After launching the job, we can see it running using our original <code>squeue</code> command: <syntaxhighlight lang="bash"> $ squeue --me JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 461 compute test.sh ben R 0:37 1 compute-permanent-node-285 </syntaxhighlight> We can cancel an in-progress job by running <code>scancel</code>: <syntaxhighlight lang="bash"> scancel 461 </syntaxhighlight> [https://github.com/kscalelabs/mlfab/blob/master/mlfab/task/launchers/slurm.py#L262-L309 Here is a reference] <code>sbatch</code> script for launching machine learning jobs. ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. ==== Useful Commands ==== Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] dabca405a0ee660bb3855898bd7d0bfa0dd8010c 0 16 1258 1257 2024-05-25T00:12:26Z Ben 2 /* SLURM Commands */ wikitext text/x-wiki The K-Scale Labs clusters are shared cluster for robotics research. This page contains notes on how to access the clusters. == Onboarding == To get onboarded, you should send us the public key that you want to use and maybe your preferred username. === Lambda Cluster === After being onboarded, you should receive the following information: * Your user ID (for this example, we'll use <code>stompy</code>) * The jumphost ID (for this example, we'll use <code>127.0.0.1</code>) * The cluster ID (for this example, we'll use <code>127.0.0.2</code>) To connect, you should be able to use the following command: <syntaxhighlight lang="bash"> ssh -o ProxyCommand="ssh -i ~/.ssh/id_rsa -W %h:%p stompy@127.0.0.1" stompy@127.0.0.2 -i ~/.ssh/id_rsa </syntaxhighlight> Note that <code>~/.ssh/id_rsa</code> should point to your private key file. Alternatively, you can add the following to your SSH config file, which should allow you to connect directly, Use your favorite editor to open the ssh config file (normally located at <code>~/.ssh/config</code> for Ubuntu) and paste the text: <syntaxhighlight lang="text"> Host jumphost User stompy Hostname 127.0.0.1 IdentityFile ~/.ssh/id_rsa Host cluster User stompy Hostname 127.0.0.2 ProxyJump jumphost IdentityFile ~/.ssh/id_rsa </syntaxhighlight> After setting this up, you can use the command <code>ssh cluster</code> to directly connect. You can also access via VS Code. Tutorial of using <code>ssh</code> in VS Code is [https://code.visualstudio.com/docs/remote/ssh-tutorial here]. Please inform us if you have any issues! === Notes === * You may need to restart <code>ssh</code> to get it working. * You may be sharing your part of the cluster with other users. If so, it is a good idea to avoid using all the GPUs. If you're training models in PyTorch, you can do this using the <code>CUDA_VISIBLE_DEVICES</code> command. * You should avoid storing data files and model checkpoints to your root directory. Instead, use the <code>/ephemeral</code> directory. Your home directory should come with a symlink to a subdirectory which you have write access to. === Andromeda Cluster === The Andromeda cluster is a different cluster which uses Slurm for job management. Authentication is different from the Lambda cluster - Ben will provide instructions directly. Don't do anything computationally expensive on the main node or you will crash it for everyone. Instead, when you need to run some experiments, reserve a GPU (see below). ==== SLURM Commands ==== Show all currently running jobs: <syntaxhighlight lang="bash"> squeue </syntaxhighlight> Show your own running jobs: <syntaxhighlight lang="bash"> squeue --me </syntaxhighlight> Show the available partitions on the cluster: <syntaxhighlight lang="bash"> sinfo </syntaxhighlight> You'll see something like this: <syntaxhighlight lang="bash"> $ sinfo PARTITION AVAIL TIMELIMIT NODES STATE NODELIST compute* up infinite 8 idle compute-permanent-node-[68,285,493,580,625-626,749,801] </syntaxhighlight> This means: * There is one compute node type, called <code>compute</code> * There are 8 nodes of that type, all currently in <code>idle</code> state * The node names are things like <code>compute-permanent-node-68</code> To launch a job, use [https://slurm.schedmd.com/srun.html srun] or [https://slurm.schedmd.com/sbatch.html sbatch]. * '''srun''' runs a command directly with the requested resources * '''sbatch''' queues the job to run when resources become available For example, suppose I have the following Shell script: <syntaxhighlight lang="bash"> #!/bin/bash echo "Hello, world!" nvidia-smi </syntaxhighlight> I can use <code>srun</code> to run this script with the following result: <syntaxhighlight lang="bash"> $ srun --gpus 8 ./test.sh Hello, world! Sat May 25 00:02:23 2024 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 550.54.15 Driver Version: 550.54.15 CUDA Version: 12.4 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| ... truncated </syntaxhighlight> Alternatively, I can queue the job using <code>sbatch</code>, which gives me the following result: <syntaxhighlight lang="bash"> $ sbatch --gpus 16 test.sh Submitted batch job 461 </syntaxhighlight> We can specify <code>sbatch</code> options inside our shell script instead using the following syntax: <syntaxhighlight lang="bash"> #!/bin/bash #SBATCH --gpus 16 echo "Hello, world!" </syntaxhighlight> After launching the job, we can see it running using our original <code>squeue</code> command: <syntaxhighlight lang="bash"> $ squeue --me JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 461 compute test.sh ben R 0:37 1 compute-permanent-node-285 </syntaxhighlight> We can cancel an in-progress job by running <code>scancel</code>: <syntaxhighlight lang="bash"> scancel 461 </syntaxhighlight> [https://github.com/kscalelabs/mlfab/blob/master/mlfab/task/launchers/slurm.py#L262-L309 Here is a reference] <code>sbatch</code> script for launching machine learning jobs. ==== Reserving a GPU ==== Here is a script you can use for getting an interactive node through Slurm. <syntaxhighlight lang="bash"> gpunode () { local job_id=$(squeue -u $USER -h -t R -o %i -n gpunode) if [[ -n $job_id ]] then echo "Attaching to job ID $job_id" srun --jobid=$job_id --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --pty $SLURM_XPUNODE_SHELL return 0 fi echo "Creating new job" srun --partition=$SLURM_GPUNODE_PARTITION --gpus=$SLURM_GPUNODE_NUM_GPUS --cpus-per-gpu=$SLURM_GPUNODE_CPUS_PER_GPU --interactive --job-name=gpunode --pty $SLURM_XPUNODE_SHELL } </syntaxhighlight> Example env vars: <syntaxhighlight lang="bash"> export SLURM_GPUNODE_PARTITION='compute' export SLURM_GPUNODE_NUM_GPUS=1 export SLURM_GPUNODE_CPUS_PER_GPU=4 export SLURM_XPUNODE_SHELL='/bin/bash' </syntaxhighlight> Integrate the example script into your shell then run <code>gpunode</code>. You can see partition options by running <code>sinfo</code>. You might get an error like this: <code>groups: cannot find name for group ID 1506</code>. But things should still run fine. Check with <code>nvidia-smi</code>. ==== Useful Commands ==== Set a node state back to normal: <syntaxhighlight lang="bash"> sudo scontrol update nodename='nodename' state=resume </syntaxhighlight> [[Category:K-Scale]] dae4374908d27b9303810d3d3b3ac9e5027cdcbf 0 282 1259 1254 2024-05-25T00:15:52Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" line> Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Loss Function === d38fd3ff4a7d1b56ed03f204961a783dbae32b22 1260 1259 2024-05-25T00:16:07Z Allen12 15 /* Overview */ wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Loss Function === a40964e9395dff252905217752aa69ef62e1b255 1261 1260 2024-05-25T00:59:49Z Allen12 15 /* Loss Function */ wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. d242e14d5f6352801c6059a5ad09af3b2de02b6c 1263 1261 2024-05-25T23:08:28Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau ~ \pi_\theta}[R(\tau)]</math> === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent b5ac0d05edd220f335fa2b7eb844a9bcb50ad83a 1264 1263 2024-05-25T23:12:04Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent 2eb997c1d926cd06a73be4efdac8e6eb636b8f4f 1265 1264 2024-05-25T23:12:37Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math> === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent bea94d920a8d64123f1256d47ddf030d3de807a5 1266 1265 2024-05-25T23:29:57Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>. === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent 90bbbd28264cae3c17a37242d6a4a931e3135c0a 1267 1266 2024-05-25T23:31:13Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>. === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent 18dd62b69ef6c8b85563989e3006e5f9b35fd255 1268 1267 2024-05-25T23:35:27Z Allen12 15 /* Objective Function */ wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. The important step here is called the Log Derivative Trick. ==Log Derivative Trick== <math>\nabla === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent 7867cdc7fbcc918644f1be8b23f8540593b38545 1269 1268 2024-05-26T00:08:48Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. The important step here is called the Log Derivative Trick. ====Log Derivative Trick==== Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent adf057324ba0ba49a5f3d5806c8e93b6c418d73a 1270 1269 2024-05-26T00:09:06Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. The important step here is called the Log Derivative Trick. ====Log Derivative Trick==== Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent f1735feff7c9680f5f7fe19b0be6c986676de3cf 1271 1270 2024-05-26T00:30:07Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. The next step here is what's called the Log Derivative Trick. ====Log Derivative Trick==== Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent 5d2b6e16b84caba2ca446308ac8df6f38f3e38f6 1272 1271 2024-05-26T00:32:41Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math> === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent f261c3bcf415668b559740287666988d1d664d2e 1273 1272 2024-05-26T00:35:11Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. Finally, using the definition of expectation again, we have <math> \nabla_\theta J(\theta) = E_{\tau \sim \pi_\theta} \left[\sum_{t=0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \right] === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent 7c7c9d881c1913efab52616f0e0aea083706d201 1274 1273 2024-05-26T00:35:24Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. Finally, using the definition of expectation again, we have <math> \nabla_\theta J(\theta) = E_{\tau \sim \pi_\theta} \left[\sum_{t=0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \right] </math> === Loss Function === The goal of REINFORCE is to optimize the expected cumulative reward. We do so using gradient descent f28b03b5b6be5159a2a62c4e9655bf6a54854d10 1275 1274 2024-05-26T00:46:17Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. Finally, using the definition of expectation again, we have <math> \nabla_\theta J(\theta) = E_{\tau \sim \pi_\theta} \left[\sum_{t=0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \right] </math> === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occuring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression \sum_\tau P(\tau | \theta) R \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t). d609021eee34d80461bd431f0ba4563d11d3b0fd 1276 1275 2024-05-26T00:46:36Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. Finally, using the definition of expectation again, we have <math> \nabla_\theta J(\theta) = E_{\tau \sim \pi_\theta} \left[\sum_{t=0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \right] </math> === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occuring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. 7683622b4988ed67dce1e1185dd66e73380e4758 1277 1276 2024-05-26T00:52:02Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. Finally, using the definition of expectation again, we have <math> \nabla_\theta J(\theta) = E_{\tau \sim \pi_\theta} \left[\sum_{t=0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \right] </math> === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have \nabla_theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right]. Using the formula for discounted reward, we have our final formula E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] 61412ac5a1624729e555cdcf77402f64cc6f0031 1278 1277 2024-05-26T00:52:28Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. Finally, using the definition of expectation again, we have <math> \nabla_\theta J(\theta) = E_{\tau \sim \pi_\theta} \left[\sum_{t=0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \right] </math> === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. df864b9df362cb671573f236c630dba167b05ed3 1279 1278 2024-05-26T00:53:11Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. f46b79ce2ac6a396da4734d914443efd360dff61 1280 1279 2024-05-26T01:13:05Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_\theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. 222d420eb14650bb883fd64580f21825217b3725 1281 1280 2024-05-26T01:17:05Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_\theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. This is why our loss is equal to -\sum_{t = 0}^T \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right e23583aa8b2f26aef53c64e55ed7c7c0c41f7648 1282 1281 2024-05-26T01:17:53Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_\theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. This is why our loss is equal to <math>-\sum_{t = 0}^T \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right</math>. d62c0c465dcd8ea55dfdef86b6478c1ccf274c07 1283 1282 2024-05-26T01:19:32Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_\theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. This is why our loss is equal to <math> -\sum_{t = 0}^T \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t </math>, since using the chain rule to take its derivative gives us the formula for the gradient.. 79c56fcbefe674013070da143ef1f9686ce4cf9d 1284 1283 2024-05-26T01:22:57Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities. We get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_\theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. This is why our loss is equal to <math> -\sum_{t = 0}^T \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t </math>, since using the chain rule to take its derivative gives us the formula for the gradient for our backwards pass (see Dennis' Optimization Notes). 70676ee0b6d291bfa62551204029135a79c52b7c 1285 1284 2024-05-26T01:23:18Z Allen12 15 wikitext text/x-wiki Allen's REINFORCE notes === Links === * [http://www.incompleteideas.net/book/RLbook2020.pdf RLbook2020] * [https://samuelebolotta.medium.com/2-deep-reinforcement-learning-policy-gradients-5a416a99700a Deep RL: Policy Gradients] [[Category:Reinforcement Learning]] === Motivation === Recall that the objective of Reinforcement Learning is to find an optimal policy <math> \pi^* </math> which we encode in a neural network with parameters <math>\theta^*</math>. <math> \pi_\theta </math> is a mapping from observations to actions. These optimal parameters are defined as <math>\theta^* = \text{argmax}_\theta E_{\tau \sim p_\theta(\tau)} \left[ \sum_t r(s_t, a_t) \right] </math>. Let's unpack what this means. To phrase it in english, this is basically saying that the optimal policy is one such that the expected value of the total reward over following a trajectory (<math> \tau </math>) determined by the policy is the highest over all policies. === Overview === ‎<syntaxhighlight lang="bash" > Initialize neural network with input dimensions = observation dimensions and output dimensions = action dimensions For each episode: While not terminated: Get observation from environment Use policy network to map observation to action distribution Randomly sample one action from action distribution Compute logarithmic probability of that action occurring Step environment using action and store reward Calculate loss over entire trajectory as function of probabilities and rewards Recall loss functions are differentiable with respect to each parameter - thus, calculate how changes in parameters correlate with changes in the loss Based on the loss, use a gradient descent policy to update weights </syntaxhighlight> === Objective Function === The goal of reinforcement learning is to maximize the expected reward over the entire episode. We use <math>R(\tau)</math> to denote the total reward over some trajectory <math>\tau</math> defined by our policy. Thus we want to maximize <math>E_{\tau \sim \pi_\theta}[R(\tau)]</math>. We can use the definition of expected value to expand this as <math>\sum_\tau P(\tau | \theta) R (\tau)</math>, where the probability of a given trajectory occurring can further be expressed as <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Now we want to find the gradient of <math> J (\theta) </math>, namely <math>\nabla_\theta \sum_\tau P(\tau | \theta) R(\tau) </math>. Since the reward function isn't a dependent on the parameters. We can rearrange: <math> \sum_\tau R(\tau) \nabla_\theta P(\tau | \theta) </math>. The next step here is what's called the Log Derivative Trick. Suppose we'd like to find <math>\nabla_{x_1}\log(f(x_1, x_2, x_3, ...))</math>. By the chain rule this is equal to <math>\frac{\nabla_{x_1}f(x_1, x_2, x_3 ...)}{f(x_1, x_2, x_3 ...)}</math>. Thus, by rearranging, we can take the gradient of any function with respect to some variable as <math>\nabla_{x_1}f(x_1, x_2, x_3, ...)= f(x_1, x_2, x_3,...)\nabla_{x_1}\log(f(x_1, x_2, x_3, ...)</math>. Thus, using this idea, we can rewrite our gradient as <math> \sum_\tau R(\tau) p(\tau | \theta) \nabla_\theta \log P(\tau | \theta) </math>. === Loss Computation === It is tricky for us to give our policy the notion of "total" reward and "total" probability. Thus, we desire to change these values parameterized by <math> \tau </math> to instead be parameterized by t. That is, instead of examining the behavior of the entire episode, we want to create a summation over timesteps. We know that <math> R(\tau) </math> is the total reward over all timesteps. Thus, we can rewrite the <math> R(\tau) </math> component at some timestep t as <math> \gamma^{T - t}r_t </math>, where gamma is our discount factor. Further, we recall that the probability of the trajectory occurring given the policy is <math> P(\tau | \theta) = P(s_0) \prod^T_{t=0} \pi_\theta(a_t | s_t) P(s_{t + 1} | s_t, a_t) </math>. Since the probabilities of <math> P(s_0) </math> and <math> P(s_{t+a} | s_t, a,t) </math> are determined by the environment and independent of the policy, their gradient is zero. Recognizing this, and further recognizing that multiplication of probabilities in log space is equal to the sum of the logarithm of each of the probabilities, we get our final gradient expression <math> \sum_\tau P(\tau | \theta) R( \tau) \sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) </math>. Rewriting this into an expectation, we have <math> \nabla_\theta J (\theta) = E_{\tau \sim \pi_\theta}\left[R(\tau)\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t)\right] </math>. Using the formula for discounted reward, we have our final formula <math> E_{\tau \sim \pi_\theta}\left[\sum_{t = 0}^T \nabla_\theta \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t \right] </math>. This is why our loss is equal to <math> -\sum_{t = 0}^T \log \pi_\theta (a_t | s_t) \gamma^{T - t}r_t </math>, since using the chain rule to take its derivative gives us the formula for the gradient for our backwards pass (see Dennis' Optimization Notes). 68c7bebc36ece8d53de937abeb18cecbcb18cd3f 0 284 1262 2024-05-25T04:57:12Z Dennisc 27 Initial notes on optimization wikitext text/x-wiki Notes of various riffs on Gradient Descent from a perspective of neural networks. [[Category: Gradient Descent]] <span id="a-review-of-standard-gradient-descent"></span> == A review of standard Gradient Descent == The goal of Gradient Descent is to minimize a loss function <math display="inline">L</math>. To be more specific, if <math display="inline">L : \mathbb R^n \to \mathbb R</math> is a differentiable multivariate function, we want to find the vector <math display="inline">w</math> that minimizes <math display="inline">L(w)</math>. Given an initial vector <math display="inline">w_0</math>, we want to “move” in the direction <math display="inline">\Delta w</math> where <math display="inline">L(w_0) - L(w_0 + \Delta w)</math> is minimized (suppose the magnitude of <math display="inline">\Delta w</math> is fixed). By Cauchy’s Inequality, this is precisely when <math display="inline">\Delta w</math> is in the direction of <math display="inline">-\nabla L(w_0)</math>. So given some <math display="inline">w_n</math>, we want to update in the direction of <math display="inline">-\alpha \nabla L(w_n)</math>. This motivates setting <math display="inline">w_{n+1} = w_n - \alpha \nabla L(w_n)</math>, where <math display="inline">\alpha</math> is a scalar factor. We call <math display="inline">\alpha</math> the “learning rate” because it affects how fast the series <math display="inline">w_n</math> converges to the optimum. The main trouble in machine learning is to tweak the <math display="inline">\alpha</math> to what “works best” in ensuring convergence, and that is one of the considerations that the remaining algorithms try to address. <span id="stochastic-gradient-descent"></span> == Stochastic Gradient Descent == In practice we don’t actually know the “true gradient”. So instead we take some datasets, say datasets <math display="inline">1</math> through <math display="inline">n</math>, and for dataset <math display="inline">i</math> we derive an estimated gradient <math display="inline">\nabla L_i</math>. Then we may estimate <math display="inline">\nabla L</math> as <math display="block">\frac{\nabla L_1 + \cdots + \nabla L_n}{n}.</math> If it is easy to compute <math display="inline">\nabla L_i(w)</math> in general then we are golden: this is the best estimate of <math display="inline">L</math> we can get. But what if <math display="inline">\nabla L_i</math> are computationally expensive to compute? Then there is a tradeoff between variance and computational cost when evaluating our estimate of <math display="inline">\nabla L</math>. A very low-cost (but low-accuracy) way to estimate <math display="inline">\nabla L</math> is just via <math display="inline">\nabla L_1</math> (or any other <math display="inline">\nabla L_i</math>). But this is obviously problematic: we aren’t even using most of our data! A better balance can be struck as follows: to evaluate <math display="inline">\nabla L(w_n)</math>, select <math display="inline">k</math> functions at random from <math display="inline">\{\nabla L_1, \ldots, \nabla L_n\}</math>. Then estimate <math display="inline">\nabla L</math> as the average of those <math display="inline">k</math> functions ''only at that step''. <span id="riffs-on-stochastic-gradient-descent"></span> == Riffs on stochastic gradient descent == <span id="momentum"></span> === Momentum === See also [https://distill.pub/2017/momentum/ “Momentum” on Distill]. In typical stochastic gradient descent, the next step we take is based solely on the gradient at the current point. This completely ignores the past gradients. However, many times it makes sense to take the past gradients into account. Of course, if we are at <math display="inline">w_{100}</math>, we should care about <math display="inline">\nabla L(w_{99})</math> much heavier than <math display="inline">\nabla L(w_1)</math>. So we should weight <math display="inline">\nabla L(w_{99})</math> much more than <math display="inline">\nabla L(w_1)</math>. The simplest way is to weight it with a geometric approach. So when we iterate, instead of taking <math display="inline">w_{n+1}</math> to satisfy <math display="block">w_{n+1} - w_n = -\alpha \nabla L(w_n)</math> like in standard gradient descent, we instead want to take <math display="inline">w_{n+1}</math> to satisfy <math display="block">w_{n+1} - w_n = -\alpha \nabla L(w_n) - \beta \alpha \nabla L(w_{n-1}) - \cdots - \beta^n \nabla L(w_0).</math> But this raises a concern: are we really going to be storing all of these terms, especially as <math display="inline">n</math> grows? Fortunately, we do not need to. For we may notice that <math display="block">w_{n+1} - w_n = -\alpha \nabla L(w_n) - \beta (\alpha \nabla L(w_{n-1}) - \cdots - \beta^{n-1} L(w_0)) = -\alpha \nabla L(w_n) - \beta(w_n - w_{n-1}).</math> To put it another way, if we write <math display="inline">w_n - w_{n-1} = \Delta w_n</math>, i.e. how much <math display="inline">w_n</math> differs from <math display="inline">w_{n-1}</math> by, we may rewrite this equation as <math display="block">\Delta w_{n+1} = -\alpha \nabla L(w_n) + \beta \Delta w_n.</math> Some of the benefits of using a momentum based approach: * most importantly, ''it can dramatically speed up convergence to a local minimum''. * it makes convergence more likely in general * escaping local minima/saddles/plateaus (its importance is possibly contested? See [https://www.reddit.com/r/MachineLearning/comments/dqbp9g/d_momentum_methods_helps_to_escape_local_minima/ this reddit thread]) <span id="rmsprop"></span> === RMSProp === Gradient descent also often has diminishing learning rates. In order to counter this, we very broadly want to - track the past learning rates, - and if they have been low, multiply <math display="inline">\Delta w_{n+1}</math> by a scalar to increase the learning rate. - (As a side effect, if our past learning rates are quite high, we will tamper the learning rates.) While performing our gradient descent to get <math display="inline">w_n \to w_{n+1}</math>, we create and store an auxillary parameter <math display="inline">v_{n+1}</math> as follows: <math display="block">v_{n+1} = \beta v_n + (1 - \beta) \nabla L(w)^2</math> and define <math display="block">w_{n+1} = w_n - \frac{\alpha}{\sqrt{v_n} + \epsilon} L(w),</math> where <math display="inline">\alpha</math> as usual is the learning rate, <math display="inline">\beta</math> is the decay rate of <math display="inline">v_n</math>, and <math display="inline">\epsilon</math> is a constant that also needs to be fine-tuned. We include the constant term of <math display="inline">\epsilon</math> in order to ensure that the sequence <math display="inline">w_n</math> actually converges and to ensure numerical stability. If we are near the minimum, then <math display="inline">v_n</math> will be quite small, meaning the denominator <math display="inline">\sqrt{v_n} + \epsilon</math> will essentially just become <math display="inline">\sqrt{v_n}</math>. But because <math display="inline">w</math> will converge when <math display="inline">L(w)</math> is just multiplied by a constant (this is the underlying assumption of standard gradient descent, after all), we will achieve convergence when near a minimum. Side note: in order to get RMSProp to interoperate with stochastic gradient descent, we instead compute the sequence <math display="inline">v_n</math> for each approximated loss function <math display="inline">L_i</math>. <span id="adam"></span> === Adam === Adam ('''Ada'''ptive '''M'''oment Estimation) is a gradient descent modification that combines Momentum and RMSProp. We create two auxillary variables while iterating <math display="inline">w_n</math> (where <math display="inline">\alpha</math> is the learning rate, <math display="inline">\beta_1</math> and <math display="inline">\beta_2</math> are decay parameters that need to be fine-tuned, and <math display="inline">\epsilon</math> is a parameter serving the same purpose as in RMSProp): <math display="block">m_{n+1} = \beta_1 m_n + (1 - \beta_1) \nabla L(w_n)</math> <math display="block">v_{n+1} = \beta_2 v_n + (1 - \beta_2) \nabla L(w_n)^2.</math> For notational convenience, we will define <math display="block">\widehat{m}_n = \frac{m_n}{1 - \beta_1^n}</math> <math display="block">\widehat{v}_n = \frac{v_n}{1 - \beta_2^n}.</math> Then our update function to get <math display="inline">w_{n+1}</math> is <math display="block">w_{n+1} = w_n - \alpha \frac{\widehat{m}_n}{\sqrt{\widehat{v}_w} + \epsilon}.</math> It is worth noting that though this formula does not explicitly include <math display="inline">\nabla L(w_n)</math>, it is accounted for in the <math display="inline">\widehat{m}_n</math> term through <math display="inline">m_n</math>. b67258a47979d4df1cb09c4b7e947c5b973bd6b7 0 1 1286 1121 2024-05-26T19:03:51Z Vrtnis 21 /* Added iCub*/ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Haier]] | |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 134202c3758f70881bbe1644662c87d64d5a39f3 0 285 1287 2024-05-26T19:09:36Z Vrtnis 21 Created page with "== Instituto Italiano di Tecnologia (IIT) == The Instituto Italiano di Tecnologia (IIT) is a leading research institution in Italy, dedicated to advancing technology and inno..." wikitext text/x-wiki == Instituto Italiano di Tecnologia (IIT) == The Instituto Italiano di Tecnologia (IIT) is a leading research institution in Italy, dedicated to advancing technology and innovation. It plays a pivotal role in the development of cutting-edge robotics and other technological advancements. == iCub Crew == The [[iCub]] crew consists of international scientists and engineers with a passion for humanoid robotics. They have delivered one of the most remarkable robotics projects, driven by their dedication and enthusiasm. === Early Developments === Initially, the team designed the mechanics, electronics, firmware, and software infrastructure of the iCub. They subsequently implemented the first examples of motor control, including inverse kinematics and force control, as well as vision libraries, simple trackers, object recognition, attention systems, and central pattern generators (CPGs) for crawling. They also applied learning methods to dynamics. Recently, the IIT team has consolidated their research along several key lines, including whole-body control (e.g., locomotion), vision and touch (both traditional and neuromorphic), grasping and manipulation, and speech and human-robot interaction. Their goal is to enhance iCub's interaction skills. === Development of R1 === The same team that developed the iCub has also created the new R1 robot. The R1 uses almost the same electronics and software as the iCub but features brand-new mechanics. c1c4ad31a19fc3ddc7e9e22890a0e3176c79da9b 0 286 1288 2024-05-26T19:12:12Z Vrtnis 21 Created page with "The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research..." wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. 4b9deaacf12f239540e584310c731511801917c2 1289 1288 2024-05-26T19:15:48Z Vrtnis 21 wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=znF1-S9JmzI | cost = Approximately €250,000 }} Introduced by IIT, iCub is designed to understand and respond to human interaction. It is utilized extensively in research settings to explore motor control, vision, touch, grasping, manipulation, speech, and human-robot interaction. == References == [https://www.iit.it/research/lines/icub IIT official website on iCub] [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Instituto Italiano di Tecnologia]] 37f2278b6cb886df35886634b60449293da05d03 1290 1289 2024-05-26T19:18:37Z Vrtnis 21 wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} Introduced by IIT, iCub is designed to understand and respond to human interaction. It is utilized extensively in research settings to explore motor control, vision, touch, grasping, manipulation, speech, and human-robot interaction. == References == [https://www.iit.it/research/lines/icub IIT official website on iCub] [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Instituto Italiano di Tecnologia]] 87909d725ea741840b79783181f1eaaacdac8a39 1291 1290 2024-05-26T19:24:51Z Vrtnis 21 /*Add general specs*/ wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} == References == [https://www.iit.it/research/lines/icub IIT official website on iCub] [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Instituto Italiano di Tecnologia]] 5ef89adec92b5fd70c5c64e8eea4c94a984e77b4 1295 1291 2024-05-26T19:31:56Z Vrtnis 21 wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} {| class="wikitable" ! Sensor type !! Number !! Notes |- | Cameras || 2 || Mounted in the eyes (see above), Pointgrey Dragonfly 2 (640x480) |- | Microphones || 2 || SoundMan High quality Stereo Omnidirectional microphone, -46 dB, 10V, 20....20 000 Hz +/- 3dB |- | Inertial sensors || 3+3 || Three axis gyroscopes + three axis accelerometers + three axis geomagnetic sensor based on BOSCH BNO055 chip, mounted in the head. (100Hz) |- | Joint sensors || For each large joint || Absolute magnetic encoder (12bit resolution @1kHz) at the joint, high-resolution incremental encoder at the motor side, hall-effect sensors for commutation (brushless motors only) |} == References == [https://www.iit.it/research/lines/icub IIT official website on iCub] [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Instituto Italiano di Tecnologia]] 6ec581afbc756367f85afea72b0676239d937896 1296 1295 2024-05-26T19:33:42Z Vrtnis 21 /* General Specifications */ wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} == Sensors == {| class="wikitable" ! Sensor type !! Number !! Notes |- | Cameras || 2 || Mounted in the eyes (see above), Pointgrey Dragonfly 2 (640x480) |- | Microphones || 2 || SoundMan High quality Stereo Omnidirectional microphone, -46 dB, 10V, 20....20 000 Hz +/- 3dB |- | Inertial sensors || 3+3 || Three axis gyroscopes + three axis accelerometers + three axis geomagnetic sensor based on BOSCH BNO055 chip, mounted in the head. (100Hz) |- | Joint sensors || For each large joint || Absolute magnetic encoder (12bit resolution @1kHz) at the joint, high-resolution incremental encoder at the motor side, hall-effect sensors for commutation (brushless motors only) |- | Joint sensors || For each small joint || Absolute magnetic encoder (except the fingers which use a custom hall-effect sensor), medium-resolution incremental encoder at the motor |- | Force/torque sensors || 6 || 6x6-axis force/torque sensors are mounted on the upper part of the arm and legs plus 2 additional sensors mounted closer to the ankle for higher precision ZMP estimation (100Hz) |- | Tactile sensors || More than 3000 (*) || Capacitive tactile sensors (8 bit resolution at 40Hz) are installed in the fingertips, palms, upper and fore-arms, chest and optionally at the legs (*). |} == References == [https://www.iit.it/research/lines/icub IIT official website on iCub] [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Instituto Italiano di Tecnologia]] e661198d4623b122cc6f140e2b93da85f7063264 1297 1296 2024-05-26T19:37:12Z Vrtnis 21 wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} == Sensors == {| class="wikitable" ! Sensor type !! Number !! Notes |- | Cameras || 2 || Mounted in the eyes (see above), Pointgrey Dragonfly 2 (640x480) |- | Microphones || 2 || SoundMan High quality Stereo Omnidirectional microphone, -46 dB, 10V, 20....20 000 Hz +/- 3dB |- | Inertial sensors || 3+3 || Three axis gyroscopes + three axis accelerometers + three axis geomagnetic sensor based on BOSCH BNO055 chip, mounted in the head. (100Hz) |- | Joint sensors || For each large joint || Absolute magnetic encoder (12bit resolution @1kHz) at the joint, high-resolution incremental encoder at the motor side, hall-effect sensors for commutation (brushless motors only) |- | Joint sensors || For each small joint || Absolute magnetic encoder (except the fingers which use a custom hall-effect sensor), medium-resolution incremental encoder at the motor |- | Force/torque sensors || 6 || 6x6-axis force/torque sensors are mounted on the upper part of the arm and legs plus 2 additional sensors mounted closer to the ankle for higher precision ZMP estimation (100Hz) |- | Tactile sensors || More than 3000 (*) || Capacitive tactile sensors (8 bit resolution at 40Hz) are installed in the fingertips, palms, upper and fore-arms, chest and optionally at the legs (*). |} {| class="wikitable" |+ Capabilities of iCub ! Task !! Description |- | Crawling || Using visual guidance with an optic marker on the floor |- | Solving complex 3D mazes || Demonstrated ability to navigate and solve intricate 3D mazes |- | Archery || Shooting arrows with a bow and learning to hit the center of the target |- | Facial expressions || Capable of expressing emotions through facial expressions |- | Force control || Utilizing proximal force/torque sensors for precise force control |- | Grasping small objects || Able to grasp and manipulate small objects such as balls and plastic bottles |- | Collision avoidance || Avoids collisions within non-static environments and can also avoid self-collision |} == References == [https://www.iit.it/research/lines/icub IIT official website on iCub] [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:Instituto Italiano di Tecnologia]] b3ba4f09cce832e6b029031716b79e15b76bc71a 0 287 1292 2024-05-26T19:27:49Z Allen12 15 Created page with "Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover..." wikitext text/x-wiki Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepislon, 1 + \varepsilon] removing incentive to go too far. d94ebb613aa8743e2846cc2c86c3abace72d6f90 1293 1292 2024-05-26T19:28:00Z Allen12 15 wikitext text/x-wiki Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepislon, 1 + \varepsilon]</math> removing incentive to go too far. 889efc6a6834ab45f1b460fbf7ba95da3c2e435e 1294 1293 2024-05-26T19:28:16Z Allen12 15 wikitext text/x-wiki Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepsilon, 1 + \varepsilon]</math> removing incentive to go too far. 4f2706d1bb897d6d4a31536371b0b4e58b81cfa4 1298 1294 2024-05-26T19:38:10Z Allen12 15 wikitext text/x-wiki === Advantage Function === <math> A(s, a) = Q(s, a) - V(s) </math>. Intuitively: extra reward we get if we take action at state compared to the mean reward at that state. We use this advantage function to tell us how good the action is - if its positive, the action is better than others at that state so we want to move in that direction, and if its negative, the action is worse than others at thtat state so we move in the opposite direction. === Motivation === Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepsilon, 1 + \varepsilon]</math> removing incentive to go too far. f179bef232fe526d26277cb4b018cefc7d4342ab 1299 1298 2024-05-26T19:47:23Z Allen12 15 wikitext text/x-wiki === Advantage Function === <math> A(s, a) = Q(s, a) - V(s) </math>. Intuitively: extra reward we get if we take action at state compared to the mean reward at that state. We use this advantage function to tell us how good the action is - if its positive, the action is better than others at that state so we want to move in that direction, and if its negative, the action is worse than others at thtat state so we move in the opposite direction. === Motivation === Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepsilon, 1 + \varepsilon]</math> removing incentive to go too far. === Ratio Function === Intuitively, if we want to measure the divergence between our old and current policies, we want some way of figuring out the difference between action-state pairs in the old and new policies. We denote this as <math> r_t(\theta) = \frac{\pi_\theta(a_t | s_t)}{\pi_{\theta_{old}}(a_t|s_t)}. A ratio greater than one indicates the action is more likely in the current policy than the old policy, and if its between 0 and 1, it indicates the opposite. b165cc3a28ddbe1d58e6f7900e1026a98e1626ab 1300 1299 2024-05-26T20:17:22Z Allen12 15 wikitext text/x-wiki === Links === Hugging face deep rl course === Advantage Function === <math> A(s, a) = Q(s, a) - V(s) </math>. Intuitively: extra reward we get if we take action at state compared to the mean reward at that state. We use this advantage function to tell us how good the action is - if its positive, the action is better than others at that state so we want to move in that direction, and if its negative, the action is worse than others at thtat state so we move in the opposite direction. Since it's often difficult and expensive to compute the Q value for all state-action pairs, we replace Q(s, a) with our sampled reward from the action. We can improve policy gradients using this objective function instead of the reward for stability. === Motivation === Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepsilon, 1 + \varepsilon]</math> removing incentive to go too far. === Ratio Function === Intuitively, if we want to measure the divergence between our old and current policies, we want some way of figuring out the difference between action-state pairs in the old and new policies. We denote this as <math> r_t(\theta) = \frac{\pi_\theta(a_t | s_t)}{\pi_{\theta_{old}}(a_t|s_t)} </math>. A ratio greater than one indicates the action is more likely in the current policy than the old policy, and if its between 0 and 1, it indicates the opposite. This ratio function replaces the log probability in the policy objective function as the way of accounting for the change in parameters. Let's step back for a moment and think about why we might want to do this. In standard policy gradients, after we use a trajectory to update our policy, the experience gained in that trajectory is now incorrect with respect to our current policy. We resolve this using importance sampling. If the actions of the old trajectory have become unlikely, the influence of that experience will be reduced. Thus, prior to clipping, our new loss function can be written in expectation form as <math> E \left[r_t(\theta)A_t\right]. === Clipping === It's easier to understand this clipping when we break it down based on why we are clipping. Let's consider some possible cases: 1a43b8c46f2b6abf482d6e0dd44ed9834f3776c6 1302 1300 2024-05-27T01:28:27Z Allen12 15 wikitext text/x-wiki === Links === Hugging face deep rl course === Advantage Function === <math> A(s, a) = Q(s, a) - V(s) </math>. Intuitively: extra reward we get if we take action at state compared to the mean reward at that state. We use this advantage function to tell us how good the action is - if its positive, the action is better than others at that state so we want to move in that direction, and if its negative, the action is worse than others at thtat state so we move in the opposite direction. Since it's often difficult and expensive to compute the Q value for all state-action pairs, we replace Q(s, a) with our sampled reward from the action. We can improve policy gradients using this objective function instead of the reward for stability. === Motivation === Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepsilon, 1 + \varepsilon]</math> removing incentive to go too far. === Ratio Function === Intuitively, if we want to measure the divergence between our old and current policies, we want some way of figuring out the difference between action-state pairs in the old and new policies. We denote this as <math> r_t(\theta) = \frac{\pi_\theta(a_t | s_t)}{\pi_{\theta_{old}}(a_t|s_t)} </math>. A ratio greater than one indicates the action is more likely in the current policy than the old policy, and if its between 0 and 1, it indicates the opposite. This ratio function replaces the log probability in the policy objective function as the way of accounting for the change in parameters. Let's step back for a moment and think about why we might want to do this. In standard policy gradients, after we use a trajectory to update our policy, the experience gained in that trajectory is now incorrect with respect to our current policy. We resolve this using importance sampling. If the actions of the old trajectory have become unlikely, the influence of that experience will be reduced. Thus, prior to clipping, our new loss function can be written in expectation form as <math> E \left[r_t(\theta)A_t\right] </math>. If we take the gradient, it actually ends up being a nearly identical equation, only with the <math> \pi_\theta(a_t | s_t) </math> being scaled by a proportional factor <math> \pi_{\theta_{old}}(a_t | s_t) </math>. === Clipping === Our clipped objective function is <math> E_t It's easier to understand this clipping when we break it down based on why we are clipping. Let's consider some possible cases: # The ratio is in the range. If the ratio is in the range, we have no reason to clip - if advantage is positive, we should encourage our policy to increase the probability of that action, and if negative, we should decrease the probability that the policy takes the action. # The ratio is lower than <math> 1 - \epsilon </math>. If the advantage is positive, we still want to increase the probability of taking that action. If the advantage is negative, then doing a policy update will decrease further the probability of taking that action, so we instead clip the gradient to 0 and don't update our weights - even though the reward here was worse, we still want to explore. # The ratio is greater than <math> 1 + \epsilon </math>. If the advantage is positive, we already have a higher probability of taking the action than in the previous policy. Thus, we don't want to update further, and get to greedy. If the advantage is negative, we clip it to <math> 1 - \epsilon </math> as usual. 352b61673581608a4876b04ac2908555ef7d3697 1303 1302 2024-05-27T01:33:43Z Allen12 15 wikitext text/x-wiki === Links === Hugging face deep rl course === Advantage Function === <math> A(s, a) = Q(s, a) - V(s) </math>. Intuitively: extra reward we get if we take action at state compared to the mean reward at that state. We use this advantage function to tell us how good the action is - if its positive, the action is better than others at that state so we want to move in that direction, and if its negative, the action is worse than others at thtat state so we move in the opposite direction. Since it's often difficult and expensive to compute the Q value for all state-action pairs, we replace Q(s, a) with our sampled reward from the action. We can improve policy gradients using this objective function instead of the reward for stability. === Motivation === Intuition: Want to avoid too large of a policy update #Smaller policy updates more likely to converge to optimal #Falling "off the cliff" might mean it's impossible to recover How we solve this: Measure how much policy changes w.r.t. previous, clip ratio to <math>[1-\varepsilon, 1 + \varepsilon]</math> removing incentive to go too far. === Ratio Function === Intuitively, if we want to measure the divergence between our old and current policies, we want some way of figuring out the difference between action-state pairs in the old and new policies. We denote this as <math> r_t(\theta) = \frac{\pi_\theta(a_t | s_t)}{\pi_{\theta_{old}}(a_t|s_t)} </math>. A ratio greater than one indicates the action is more likely in the current policy than the old policy, and if its between 0 and 1, it indicates the opposite. This ratio function replaces the log probability in the policy objective function as the way of accounting for the change in parameters. Let's step back for a moment and think about why we might want to do this. In standard policy gradients, after we use a trajectory to update our policy, the experience gained in that trajectory is now incorrect with respect to our current policy. We resolve this using importance sampling. If the actions of the old trajectory have become unlikely, the influence of that experience will be reduced. Thus, prior to clipping, our new loss function can be written in expectation form as <math> E \left[r_t(\theta)A_t\right] </math>. If we take the gradient, it actually ends up being a nearly identical equation, only with the <math> \pi_\theta(a_t | s_t) </math> being scaled by a proportional factor <math> \pi_{\theta_{old}}(a_t | s_t) </math>. === Clipping === Our clipped objective function is <math> E_t </math> It's easier to understand this clipping when we break it down based on why we are clipping. Let's consider some possible cases: # The ratio is in the range. If the ratio is in the range, we have no reason to clip - if advantage is positive, we should encourage our policy to increase the probability of that action, and if negative, we should decrease the probability that the policy takes the action. # The ratio is lower than <math> 1 - \epsilon </math>. If the advantage is positive, we still want to increase the probability of taking that action. If the advantage is negative, then doing a policy update will decrease further the probability of taking that action, so we instead clip the gradient to 0 and don't update our weights - even though the reward here was worse, we still want to explore. # The ratio is greater than <math> 1 + \epsilon </math>. If the advantage is positive, we already have a higher probability of taking the action than in the previous policy. Thus, we don't want to update further, and get to greedy. If the advantage is negative, we clip it to <math> 1 - \epsilon </math> as usual. 553320fbafa2af7d2922e9e43b949aa52d253932 0 235 1301 1025 2024-05-27T01:13:29Z Vedant 24 wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === * How do we connect I2C chip to the LCD screen? === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery === Questions === # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? === Notes === * Need to figure out the interface for the SPI lines using MCP2515 * Pros if successful: * Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects * Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 0f5ecaa1254dcdf0e71838ab45fb2009d0491f57 1304 1301 2024-05-27T09:02:27Z Ben 2 /* Notes */ wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === * How do we connect I2C chip to the LCD screen? === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery === Questions === # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? === Notes === * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 4f4a89052402d0ea3f5aeaa7ccc9909fd64fc1d4 0 155 1305 985 2024-05-27T18:13:54Z Dymaxion 22 /* MCP2515 Driver */ wikitext text/x-wiki The '''Controller Area Network''' (CAN) represents an essential vehicle bus standard. It was explicitly engineered to facilitate communication between microcontrollers and devices without resorting to a central computer. The primary purpose of CAN is to enhance vehicle interconnectivity and foster swift data exchange between various systems within the vehicle. == MCP2515 == The '''MCP2515''' is an integrated circuit produced by Microchip Technology, constructed to function as a stand-alone CAN controller. It exhibits compatibility with an SPI (Serial Peripheral Interface) which makes it versatile in various applications. Although its primary use is in automotive industries, it is also used in a variety of other control applications.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515 bridges the connection between the CAN protocol and the SPI protocol by receiving CAN messages and translating them into SPI data, allowing the microcontroller to interpret the information. Similarly, it transforms SPI data into CAN messages for transmission.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> The MCP2515's operational details include its capacity to support CAN 2.0A and B, drawing attention to its alignment with established CAN standards and equipping it for both basic and extended frame format usage.<ref>MCP2515 Stand-Alone CAN Controller with SPI Interface. Microchip Technology. [https://www.microchip.com/wwwproducts/en/en010406 Official datasheet].</ref> == Applications == The [https://python-can.readthedocs.io/en/stable/ python-can] library provides Controller Area Network support for Python, providing common abstractions to different hardware devices, and a suite of utilities for sending and receiving messages on a CAN bus. There is an example of an installation on [[Jetson_Orin]] here. === Pi4 === Raspberry Pi offers an easy to deploy 2-Channel Isolated CAN Bus Expansion HAT which allows to quickly integrate it to the peripheral devices. See the [https://www.waveshare.com/wiki/2-CH_CAN_HAT tutorial] for more information === Arduino === Arduino has a good support of the MCPs with many implementations of the [https://github.com/Seeed-Studio/Seeed_Arduino_CAN drivers] === MCP2515 Driver === By default the CAN bus node is supposed to acknowledge every message on the bus whether or not that node is interested in the message. However, the interference on the network can drop some bits during the communication. In the standard mode, the node would not only continuously try to re-send the unacknowledged messages, but also after a short period it would start sending error frames and then eventually go to bus-off mode and stop. This causes sever issues when the CAN network works with multiple motors. The controller has a [http://ww1.microchip.com/downloads/en/DeviceDoc/MCP2515-Stand-Alone-CAN-Controller-with-SPI-20001801J.pdf one-shot] setup that requires changes in the driver. === Wiring === Here is some suggested equipment for wiring a CAN bus: * Molex Mini-Fit Junior connectors ** [https://www.digikey.ca/en/products/detail/molex/0638190901/9655931 Crimper] ** [https://www.digikey.ca/en/products/detail/molex/0766500013/2115996 Connector kit] ** [https://www.aliexpress.us/item/3256805730106963.html Extraction tool] * CAN bus cable ** [https://www.digikey.ca/en/products/detail/igus/CF211-02-01-02/18724291 Cable] ** [https://www.digikey.com/en/products/detail/igus/CF891-07-02/21280679 Alternative Cable] * Heat shrink ** [https://www.amazon.com/Eventronic-Heat-Shrink-Tubing-Kit-3/dp/B0BVVMCY86 Tubing] == References == <references /> [[Category:Communication]] ffa771efe644014d3a1453d4524b54a96e99a471 6 288 1306 2024-05-27T22:45:36Z Kewang 11 wikitext text/x-wiki show the hierarchy of the folers 7e2b60613f419828d6d298d4c41475b8bc1f9e70 6 289 1307 2024-05-27T22:58:26Z Kewang 11 wikitext text/x-wiki model1 3bd6e748431d451c4dd72b968cf047aa75f6b493 6 290 1308 2024-05-27T23:01:13Z Kewang 11 wikitext text/x-wiki model2 a6a4646aa4d1de42279ca51938186ee134bb4403 6 291 1309 2024-05-27T23:09:28Z Kewang 11 wikitext text/x-wiki add_model3 059d8879a7db05c1d002aeb9a420d2f0f4855ea6 0 257 1310 1217 2024-05-27T23:09:58Z Kewang 11 wikitext text/x-wiki == Install emscripten == First, you need to install emscripten, which is a compiler toolchain for WebAssembly. === Get the emsdk repo === <code> git clone https://github.com/emscripten-core/emsdk.git </code> === Enter that directory === <code> cd emsdk </code> === Download and install the latest SDK tools === <code> ./emsdk install latest </code> === Make the "latest" SDK "active" === <code> ./emsdk activate latest </code> === Activate PATH and other environment variables === <code> source ./emsdk_env.sh </code> These variables are set for the current terminal. If you want to make it for all terminals, you can add them to any terminal profile. Here they are: The environment variables: <code> EMSDK = < path to emsdk dir > EM_CONFIG = ~/.emscripten EMSDK_NODE = < path to emsdk dir >/node/12.9.1_64bit/bin/node </code> === Now just try it! === <code> emcc </code> == Build the mujoco_wasm Binary == First git clone <code> https://github.com/zalo/mujoco_wasm </code> Next, you'll build the MuJoCo WebAssembly binary. <syntaxhighlight lang="bash"> mkdir build cd build emcmake cmake .. make </syntaxhighlight> [[File:Carbon (1).png|800px|thumb|none|emcmake cmake ..]] [[File:Carbon (2).png|400px|thumb|none|make]] '''Tip:''' If you get an error with "undefined symbol: saveSetjmp/testSetjmp" at the build step, revert to: <code> ./emsdk install 3.1.56 && ./emsdk activate 3.1.56 && source ./emsdk_env.sh </code> == Running in Browser == Run this in your mujoco folder to start a server. <code> python -m http.server 8000 </code> Then navigate to: <code> http://localhost:8000/index.html </code> [[File:Wasm screenshot13-40-40.png|800px|thumb|none|MuJoCo running in browser]] == Running in Cloud/Cluster and Viewing on Local Machine == Add extra parameter to your ssh command: <code> ssh -L 8000:127.0.0.1:8000 my_name@my_cluster_ip </code> Then you can open it on the browser on your local machine! == Adding New Models == All the models are stored in the folder examples/scenes, as seen below: [[File:Example folder.png|400px|thumb|none]] You can add your own model XML and meshes here. For example, here we add the stompy folder. After adding the model files, run <code> python generate_index.py </code> to update file indexes. Then copy all the content in <code>index.json</code> to <code> mujocoUtils.js</code>, as shown below: [[File:Add model1.png|400px|thumb|none]] In the end, again at file <code> mujocoUtils.js</code>, add the name and scene file [[File:Add model2.png|600px|thumb|none]] Then reload again and you can see new models have been added: [[File:Add model3.png|800px|thumb|none]] 1bf74441f94674f9fcb1230ebbc621707daf865c 0 235 1311 1304 2024-05-28T01:16:11Z Vedant 24 wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === == PCF2518 == * How do we connect I2C chip to the LCD screen? === Battery Charger === === MCP 2515 === === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery === Questions === # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? === Notes === * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 090729d60af80d647261efe8c0e1e191647705d4 1312 1311 2024-05-28T01:17:04Z Vedant 24 /* PCF2518 */ wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === === PCF2518 === * How do we connect I2C chip to the LCD screen? === Battery Charger === === MCP 2515 === === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery === Questions === # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? === Notes === * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. da951a035c608beab5223c6623408aeb88b09ead 1313 1312 2024-05-28T01:24:25Z Vedant 24 wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === === PCF2518 === * How do we connect I2C chip to the LCD screen? === Battery Charger === === MCP 2515 === === Github access === === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery === Questions === # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? === Notes === * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 60d263fdcf6ed36fd547dca8b9d59c3a9988bbd6 1314 1313 2024-05-28T07:28:51Z Vedant 24 wikitext text/x-wiki === MCU === * STM32 ** Programming mode button (boot / reset) Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs === USB === * adsfasdfasdf === Buttons === * Button debouncer ** TODO: Add the IC name === LCD === === PCF2518 === * How do we connect I2C chip to the LCD screen? === Battery Charger === === MCP 2515 === === Github access === === CAN === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? === Voltage === * Load sharing between USB and battery === Questions === # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? === Notes === * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 9765b636c42af3f8c478643e1674ed0264e79f11 1315 1314 2024-05-28T07:34:24Z Vedant 24 wikitext text/x-wiki = Features = # USB-C + Battery compatibility # Able === Voltage === * Load sharing between USB and battery = Central Components = == MCU == = STM32 = Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs == Buttons + Debouncer == * Button Debouncer ** TODO: Add the IC name == LCD == === PCF2518 === * How do we connect I2C chip to the LCD screen? == Battery Charger == === MCP 2515 === == CAN Communication == === MCP2515 === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? = Other = == Questions == # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? == Notes == * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 110bb3a215ceb3f523f356cf9b72fc31f0344a95 1316 1315 2024-05-28T07:34:54Z Vedant 24 wikitext text/x-wiki = Features = # USB-C + Battery compatibility # Able === Voltage === * Load sharing between USB and battery = Central Components = == MCU == === STM32 === Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs == Buttons + Debouncer == * Button Debouncer ** TODO: Add the IC name == LCD == === PCF2518 === * How do we connect I2C chip to the LCD screen? == Battery Charger == === MCP 2515 === == CAN Communication == === MCP2515 === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? = Other = == Questions == # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? == Notes == * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. a272a138723ba1dd3057aa25403caedaf4eb526f 1317 1316 2024-05-28T07:35:47Z Vedant 24 wikitext text/x-wiki = Features = # USB-C + Battery compatibility # Able === Voltage === * Load sharing between USB and battery = Central Components = == MCU == === STM32 === Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs == Buttons + Debouncer == * Button Debouncer ** TODO: Add the IC name == LCD == === PCF2518 === * How do we connect I2C chip to the LCD screen? == Battery Charger == === MCP 2515 === == CAN Communication == === MCP2515 === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? = Design = == Schematic == == Layout == == Final Form Factor == = Other = == Questions == # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? == Notes == * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. e58654874723a9af6fa33d8c927be95534602fb2 1318 1317 2024-05-28T07:36:18Z Vedant 24 /* CAN Communication */ wikitext text/x-wiki = Features = # USB-C + Battery compatibility # Able === Voltage === * Load sharing between USB and battery = Central Components = == MCU == === STM32 === Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs == Buttons + Debouncer == * Button Debouncer ** TODO: Add the IC name == LCD == === PCF2518 === * How do we connect I2C chip to the LCD screen? == Battery Charger == === MCP 2515 === == CAN Communication == === MCP2515 === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? = Design = == Schematic == == Layout == == Final Form Factor == = Other = == Questions == # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? == Notes == * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 350c104dc359246468d54f276aa4e0cdc4e49b74 1319 1318 2024-05-28T07:36:29Z Vedant 24 /* MCP 2515 */ wikitext text/x-wiki = Features = # USB-C + Battery compatibility # Able === Voltage === * Load sharing between USB and battery = Central Components = == MCU == === STM32 === Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs == Buttons + Debouncer == * Button Debouncer ** TODO: Add the IC name == LCD == === PCF2518 === * How do we connect I2C chip to the LCD screen? == Battery Charger == === MCP 2515 === == CAN Communication == === MCP2515 === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? = Design = == Schematic == == Layout == == Final Form Factor == = Other = == Questions == # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? == Notes == * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. e26b1f9d788ebd80598bce318e126bc7133e816b 1320 1319 2024-05-28T07:36:41Z Vedant 24 /* Battery Charger */ wikitext text/x-wiki = Features = # USB-C + Battery compatibility # Able === Voltage === * Load sharing between USB and battery = Central Components = == MCU == === STM32 === Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs == Buttons + Debouncer == * Button Debouncer ** TODO: Add the IC name == LCD == === PCF2518 === * How do we connect I2C chip to the LCD screen? == Battery Charger == == CAN Communication == === MCP2515 === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? = Design = == Schematic == == Layout == == Final Form Factor == = Other = == Questions == # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? == Notes == * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. 1f6293006629245e4e0d9e037f80d5074af1b2f0 1321 1320 2024-05-28T07:38:08Z Vedant 24 wikitext text/x-wiki = Features = # USB-C + Battery compatibility # Able === Voltage === * Load sharing between USB and battery = Central Components = == MCU == === STM32 === Used: STM32F407VET6 Justification: 512 kB of flash memory, unit price at $3, powerful enough to run heavy programs == Buttons + Debouncer == * Button Debouncer ** TODO: Add the IC name == LCD == === PCF2518 === * How do we connect I2C chip to the LCD screen? == Battery Charger == == CAN Communication == === MCP2515 === * How to get supply-side VCC without having it on the bus (from supply-side GND)? * What do we do if supply-side GND is missing? = Design = == Schematic == == Layout == == Final Form Factor == = Other = == Questions == # How does the switching regulator mechanism work? # Do digital logic levels depend on source voltage? For example, let's say that a 3.3V powered device is sending signals to a 5V power device. From my understanding of the datasheet, a direct connection between the signals should not be allowed. However, why is this? What are the internal mechanisms of the pins that make it so that the logic levels need to be based off of the source voltage? # I understand that there is a way to both use the battery as a source as well as to charge it using the USB. However, wouldn't this mean not being able to put a diode within the circuit directing current in one direction? Isn't this a hazard, destroying the circuit the moment a bit of noise comes through? # Power Ground in the context of a switching regulator for our circuit to amplify 3.7 V battery to 5 V source. Do we just connect Power Ground to Battery Ground? == Notes == * Need to figure out the interface for the SPI lines using MCP2515 ** Pros if successful: *** Reduces amount of wiring for STM32, requires just one SPI bus with a bunch of easily programmable chip selects ** Cons: Need to understand the interface for the MCP2515. Need to confirm how interrupts and timing works. == Github Repository == 2e90c40af66f70ca693545f09b8242ef93b83aa5 0 292 1322 2024-05-28T07:45:02Z Vedant 24 Created page with "= Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs ca..." wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = == ESP32 Built-In == By connecting 34c7396f7f0dbb10f02fd7543e05c40e59f56b8a 1323 1322 2024-05-28T08:30:55Z Vedant 24 /* ESP32 Built-In */ wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * == Pros and Cons == === Pros === === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === * Contains both Bluetooth as well as Wifi protocols * Uses UART for Bluetooth * Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. = af2fbe84fb64a2da7a18e203cd3422a15daa3df4 1324 1323 2024-05-28T08:31:13Z Vedant 24 /* CYM4343W */ wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * == Pros and Cons == === Pros === === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === Contains both Bluetooth as well as Wifi protocols Uses UART for Bluetooth Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. = 187555a6084684454a7dfee2ec9f00f1439da486 1325 1324 2024-05-28T08:32:33Z Vedant 24 /* Pros */ wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * == Pros and Cons == === Pros === Allows for convenient, easy to mount access to wireless communication. Easy to interface with, using UART in order to communicate data. === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === Contains both Bluetooth as well as Wifi protocols Uses UART for Bluetooth Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. = 6018eefdbd9b30b0377a251623bd7bd794c7861f 1326 1325 2024-05-28T08:32:46Z Vedant 24 /* Pros and Cons */ wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * == Pros and Cons == === Pros === Allows for convenient, easy to mount access to wireless communication. Easy to interface with, using UART in order to communicate data. === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === Contains both Bluetooth as well as Wifi protocols Uses UART for Bluetooth Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with/ Not modular. == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. = 7fcc24671a00df8882f9f446e80c65ed5d4b8545 1327 1326 2024-05-28T08:32:55Z Vedant 24 /* TODO */ wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * == Pros and Cons == === Pros === Allows for convenient, easy to mount access to wireless communication. Easy to interface with, using UART in order to communicate data. === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === Contains both Bluetooth as well as Wifi protocols Uses UART for Bluetooth Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with/ Not modular. == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. c4ec65a49c5e401513b57bba0fca23e30f3de9b8 1328 1327 2024-05-28T08:34:54Z Vedant 24 /* UART Communication */ wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * == Pros and Cons == === Pros === Allows for convenient, easy to mount access to wireless communication. Easy to interface with, using UART in order to communicate data. === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === Contains both Bluetooth as well as Wifi protocols Uses UART for Bluetooth Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with/ Not modular. == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. 0dc652df614e57c69c9c7c4a58288dca10ce6a85 1329 1328 2024-05-28T08:36:23Z Vedant 24 /* Examples */ wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-WROOM-32-N4/8544298 ESP32-WROOM-32] * uses UART To both receive and send data to other modules == Pros and Cons == === Pros === Allows for convenient, easy to mount access to wireless communication. Easy to interface with, using UART in order to communicate data. === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === Contains both Bluetooth as well as Wifi protocols Uses UART for Bluetooth Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with/ Not modular. == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. 0843c162296492f539fb687b120cdb729a403530 1330 1329 2024-05-28T08:52:50Z Vedant 24 wikitext text/x-wiki = Justification = Robust, wireless communication between multiple PCB boards decreased design constraints and wiring, allowing for much more modular and adjustable designs capable. This would allow for mobility, ease of communication, reduced clutter, and scalability. Could potentially = Current Methods of Integration = = ESP32 Built-In = == Connections == === UART Communication === * Hook up UART_TX and UART_RX to STM32. === I2C Communication === === SPI Communication === == Firmware == == Examples == [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-C3-MINI-1-N4/13877574 ESP32-C3-MINI-1-N4] * ESP32 Chip adjusted specifically for the sole purpose of wireless communication. * Interfaced with using UART * [https://www.digikey.com/en/products/detail/espressif-systems/ESP32-WROOM-32-N4/8544298 ESP32-WROOM-32] * uses UART To both receive and send data to other modules == Pros and Cons == === Pros === Allows for convenient, easy to mount access to wireless communication. Easy to interface with, using UART in order to communicate data. === Cons === = Bluetooth Modules = == Component Selection == === HC_05/HC_06 === === CYM4343W === Contains both Bluetooth as well as Wifi protocols Uses UART for Bluetooth Uses SDIO for wifi communication == Pros and Cons == === Pros === * Simple implementation with supported Code === Cons === * Separate PCB Board interfaced with/ Not modular. == TODO == Look into the specific components of the Bluetooth modules to find ICS that can be easily used and interfaced with. Figure out how to interface with ESP32 software using STM32, how to send specific commands, etc. = Interfacing with Other Boards = == Network Setup == === Access Point === Set up one device as the master. Intended purpose of the CANdaddy. Host signal and allow other devices to connect using SSID and password. In the case of the ESP32, this is done through built in libraries that allow for == Communication Protocol == == Data Exchange == dfcf9693ffb8df5206e24919c5aca4e5bad9a8ff 0 293 1331 2024-05-28T17:21:27Z Ben 2 Created page with "=== Papers === * [https://distill.pub/2017/ctc/ CTC] * [https://arxiv.org/abs/1810.04805 BERT] * [https://arxiv.org/abs/2006.11477 wav2vec 2.0] * [https://arxiv.org/abs/2106...." wikitext text/x-wiki === Papers === * [https://distill.pub/2017/ctc/ CTC] * [https://arxiv.org/abs/1810.04805 BERT] * [https://arxiv.org/abs/2006.11477 wav2vec 2.0] * [https://arxiv.org/abs/2106.07447 HuBERT] * [https://speechbot.github.io/ Textless NLP project] * [https://arxiv.org/abs/2210.13438 Encodec] * [https://arxiv.org/abs/2308.16692 SpeechTokenizer] * [https://github.com/suno-ai/bark Suno Bark Model] 6db6779ee74eccd1c8ec7c5b8e1848ec99c9f73c 0 281 1332 1230 2024-05-28T19:41:00Z Vrtnis 21 /*Add ANYmal, HRP-2*/ wikitext text/x-wiki === Humanoids === === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | TIAGo || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | ANYmal || ANYbotics || URDF, SDF || [https://github.com/ANYbotics/anymal_b_simple_description URDF], [https://github.com/ANYbotics/anymal_c_simple_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-2 || Kawada Robotics || URDF || [https://github.com/start-jsk/rtmros_common/tree/master/hrp2_models URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | HRP-4 || Kawada Robotics || URDF || [https://github.com/start-jsk/rtmros_common/tree/master/hrp4_models URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 5e39cff184d6b5c0430b0f5095ce4898d35717cb 1333 1332 2024-05-28T19:41:20Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | TIAGo || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | ANYmal || ANYbotics || URDF, SDF || [https://github.com/ANYbotics/anymal_b_simple_description URDF], [https://github.com/ANYbotics/anymal_c_simple_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-2 || Kawada Robotics || URDF || [https://github.com/start-jsk/rtmros_common/tree/master/hrp2_models URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | HRP-4 || Kawada Robotics || URDF || [https://github.com/start-jsk/rtmros_common/tree/master/hrp4_models URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} c9811ef97fc6bbfd00ce0c139ce5df3c02e00335 1334 1333 2024-05-28T19:46:12Z Vrtnis 21 /*Add new robots and references*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | TIAGo || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | ANYmal || ANYbotics || URDF, SDF || [https://github.com/ANYbotics/anymal_b_simple_description URDF], [https://github.com/ANYbotics/anymal_c_simple_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-2 || Kawada Robotics || URDF || [https://github.com/start-jsk/rtmros_common/tree/master/hrp2_models URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | HRP-4 || Kawada Robotics || URDF || [https://github.com/start-jsk/rtmros_common/tree/master/hrp4_models URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 0e16ea5381eaeb2fa8f7d85ad19dbde99b3d7bfe 1335 1334 2024-05-28T19:54:30Z Vrtnis 21 /*Add Rethink Baxter*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} f354b8f98c5441b8f55769fe99fec01f17cbb3e2 1336 1335 2024-05-28T19:57:51Z Vrtnis 21 /*Add Baxter End Effector*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 0e920bc7e3de9092f18d4cbdc2384be9b8520aa8 1337 1336 2024-05-28T20:03:18Z Vrtnis 21 /* Add Pepper */ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} 892c00e87a797332d795f01ba7a19104c02a5de0 1338 1337 2024-05-28T20:05:11Z Vrtnis 21 /*Add Pepper references*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology d49ec4e30e74e2b51ecdb88dee90cb73b439022a 1339 1338 2024-05-28T20:07:16Z Vrtnis 21 /*Add Mini-Cheetah*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology a36bdae9e39c24c601c2836fbc05210e67f16a40 1340 1339 2024-05-28T20:07:58Z Vrtnis 21 /*Add Mini-Cheetah*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology fb835c3a528cd1aa34abd13f60ff79b9bbd245e8 1341 1340 2024-05-28T20:09:33Z Vrtnis 21 /* Add Thor Mang */ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology f1f863aa3936f9ca3c519e584d231cdcc96c8f75 1342 1341 2024-05-28T20:12:24Z Vrtnis 21 /* Baxter End Effectors */ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 76c3778e1bff91bc8248fd4aa0e15af72fceb87c 1343 1342 2024-05-28T20:25:59Z Vrtnis 21 /*Asimo, Sophia, Cassie*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology ef5331d02e39383d8959448b0053be84c84829e3 1344 1343 2024-05-29T00:45:40Z Vrtnis 21 /*Add Kawada Robotics*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology c1713f349ff441644e9bf91cec89f521d2498737 1347 1344 2024-05-29T04:25:41Z Vrtnis 21 /* Add NASA Valkyrie*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology f662b7edaf0ecbf806b2546efaf7c6bdd8a9b7c8 1348 1347 2024-05-29T04:26:58Z Vrtnis 21 /* Add REEM-C */ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 149f48299863f1b865c8c263ff7721ae06c0e716 0 294 1345 2024-05-29T02:46:19Z Ben 2 Created page with "{| class="wikitable" |- ! Link |- | [https://twitter.com/kscalelabs/status/1788968705378181145 2024.05.10] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17..." wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://twitter.com/kscalelabs/status/1788968705378181145 2024.05.10] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |} 2047a5c01d3fe9522dadc6b7f119a591d9f2d1b6 1346 1345 2024-05-29T03:12:28Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://twitter.com/kscalelabs/status/1788968705378181145 2024.05.10] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |} [[Category:K-Scale]] 6eba57fc29aa717215ff25a4ef681f43fb83a5db 0 1 1349 1286 2024-05-29T04:38:34Z Vrtnis 21 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 09843c8b175405fd4665cfad14c904de6b87ff8b 1350 1349 2024-05-29T04:41:02Z Vrtnis 21 /* Add Stanford Robotics Lab */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} ab75e3ff270ec6bfe8e258b5c1d0d888cc982638 1355 1350 2024-05-29T17:34:31Z Ben 2 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[Tesla]] | [[Optimus]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[WorkFar]] | [[WorkFar Syntro]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} e590bc3a348ff7bbce568e0bbf507b64d1b2b376 0 295 1351 2024-05-29T04:52:45Z Vrtnis 21 /*Add Stanford Robotics Lab*/ wikitext text/x-wiki The Stanford Robotics Lab, part of the Stanford Artificial Intelligence Laboratory (SAIL), is a research facility dedicated to the advancement of robotics technology. Located within Stanford University, the lab focuses on a wide range of research areas, including autonomous systems, human-robot interaction, robotic manipulation, and machine learning applications in robotics. The lab's interdisciplinary approach combines expertise from various fields such as computer science, electrical engineering, mechanical engineering, and cognitive sciences to develop innovative robotic solutions. Researchers at the Stanford Robotics Lab work on projects that aim to improve the capabilities and intelligence of robots, enabling them to perform complex tasks in dynamic and unstructured environments. Notable achievements of the lab include significant contributions to the development of autonomous vehicles, advanced prosthetics, and collaborative robots that can safely and efficiently work alongside humans. The lab also emphasizes real-world applications and often collaborates with industry partners to bring cutting-edge research into practical use. [https://robotics.stanford.edu Stanford Robotics Lab website] 05592a2cd3fd83d7fe2c00e8518c0e921d364eaa 0 296 1352 2024-05-29T04:55:44Z Vrtnis 21 /*Add OcenOneK*/ wikitext text/x-wiki '''OceanOneK''' is an advanced underwater humanoid robot developed by the Stanford Robotics Lab. OceanOneK is equipped with stereoscopic vision, haptic feedback, and fully articulated arms and hands, allowing operators to perform delicate and precise tasks in underwater settings. Key features of OceanOneK include: * '''Stereoscopic Vision''': Provides operators with a clear, three-dimensional view of the underwater environment, enhancing situational awareness and precision. * '''Haptic Feedback''': Allows operators to feel the force and resistance encountered by the robot's hands, providing a tactile sense of the underwater objects being manipulated. * '''Articulated Arms and Hands''': Enable the robot to perform complex tasks such as collecting samples, repairing equipment, and interacting with delicate marine life. [https://robotics.stanford.edu/oceanonek OceanOneK project page] 51b9301c2d1441827aa5017f87c91475de12074d 1353 1352 2024-05-29T05:10:09Z Vrtnis 21 /*Add infobox*/ wikitext text/x-wiki {{Infobox robot | name = OceanOneK | image = | caption = OceanOneK robot during a deep-sea mission | country = United States | year_of_creation = 2022 | developer = Stanford Robotics Lab | type = Underwater Humanoid Robot | purpose = Deep-sea exploration, Archaeology, Environmental monitoring | weight = 235 kg | height = 1.8 meters | manipulator = Fully articulated arms and hands | vision = Stereoscopic vision | feedback = Haptic feedback | control = Remote operator | website = [https://robotics.stanford.edu/oceanonek OceanOneK project page] }} '''OceanOneK''' is an advanced underwater humanoid robot developed by the Stanford Robotics Lab. OceanOneK is equipped with stereoscopic vision, haptic feedback, and fully articulated arms and hands, allowing operators to perform delicate and precise tasks in underwater settings. Key features of OceanOneK include: * '''Stereoscopic Vision''': Provides operators with a clear, three-dimensional view of the underwater environment, enhancing situational awareness and precision. * '''Haptic Feedback''': Allows operators to feel the force and resistance encountered by the robot's hands, providing a tactile sense of the underwater objects being manipulated. * '''Articulated Arms and Hands''': Enable the robot to perform complex tasks such as collecting samples, repairing equipment, and interacting with delicate marine life. [https://robotics.stanford.edu/oceanonek OceanOneK project page] 8e03d41f301eb1f2a422b67ea5c54aaa0ee82f01 1354 1353 2024-05-29T05:11:42Z Vrtnis 21 wikitext text/x-wiki {{Infobox robot | name = OceanOneK | organization = [[Stanford Robotics Lab]] | height = 1.8 meters | weight = 235 kg | purpose = Deep-sea exploration, Archaeology, Environmental monitoring | manipulator = Fully articulated arms and hands | vision = Stereoscopic vision | feedback = Haptic feedback | control = Remote operator | website = https://robotics.stanford.edu/oceanonek }} '''OceanOneK''' is an advanced underwater humanoid robot developed by the Stanford Robotics Lab. OceanOneK is equipped with stereoscopic vision, haptic feedback, and fully articulated arms and hands, allowing operators to perform delicate and precise tasks in underwater settings. Key features of OceanOneK include: * '''Stereoscopic Vision''': Provides operators with a clear, three-dimensional view of the underwater environment, enhancing situational awareness and precision. * '''Haptic Feedback''': Allows operators to feel the force and resistance encountered by the robot's hands, providing a tactile sense of the underwater objects being manipulated. * '''Articulated Arms and Hands''': Enable the robot to perform complex tasks such as collecting samples, repairing equipment, and interacting with delicate marine life. [https://robotics.stanford.edu/oceanonek OceanOneK project page] a9365161e4cd733547c238030271693c5fcc9e3f 0 297 1356 2024-05-29T17:35:21Z Ben 2 Created page with "WorkFar is a US-based company building a humanoid robot for factory automation. {{infobox company | name = WorkFar | country = USA | website_link = https://www.workfar.com/ |..." wikitext text/x-wiki WorkFar is a US-based company building a humanoid robot for factory automation. {{infobox company | name = WorkFar | country = USA | website_link = https://www.workfar.com/ | robots = [[WorkFar Syntro]] }} [[Category:Companies]] d20c9cd6994a53455dd2e36e6da000de55387886 0 298 1357 2024-05-29T17:36:30Z Ben 2 Created page with "The '''Syntro''' robot from [[WorkFar]] is designed for advanced warehouse automation. {{infobox robot | name = Syntro | organization = [[WorkFar]] | video_link = https://ww..." wikitext text/x-wiki The '''Syntro''' robot from [[WorkFar]] is designed for advanced warehouse automation. {{infobox robot | name = Syntro | organization = [[WorkFar]] | video_link = https://www.youtube.com/watch?v=suF7mEtLJvY | cost = | height = | weight = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} [[Category:Robots]] a1972a23ba3237855ac71e9086d80612bdafa18a 0 281 1358 1348 2024-05-29T22:34:38Z Vrtnis 21 /*Add Darwin OP*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ - | Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology c6a269feb47bdb0f4fc7dfc2f57ab846dfd00ec6 1359 1358 2024-05-29T22:35:42Z Vrtnis 21 /* Add Darwin OP URDF */ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology e9d62f6cb35ee61958ea80076a7fc312ee723c27 1360 1359 2024-05-29T22:43:50Z Vrtnis 21 /*Add INRIA Poppy Humanoid*/ wikitext text/x-wiki === Humanoids === {| class="wikitable" |- ! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 4bd7515b010f82daacaffc2222526372e67f7c5f 1361 1360 2024-05-29T22:46:11Z Vrtnis 21 /*Make sortable*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology c24af55b0b161b035b33da1f972040d13c7e24fd 1362 1361 2024-05-29T22:54:45Z Vrtnis 21 /*Add Kengoro and SURALP*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 12dc09409bfbb8fc97f6af0c436ae5cef9f62c4b 1371 1362 2024-05-31T05:22:25Z Vrtnis 21 /* Kengoro*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology cee333b4d2b81b44a1dfae4170f0fb7fe724f2e9 1372 1371 2024-05-31T05:25:34Z Vrtnis 21 /* ANYmal wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology eb1ee085e56e0b6e75574f3664e1b1df0b5e7753 1373 1372 2024-05-31T05:30:23Z Vrtnis 21 /*MIR */ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 562afa83e686f21d7d80858255086219f39baa71 1374 1373 2024-05-31T05:34:14Z Vrtnis 21 /*HSR and Pepper2*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology c3d2994b5bbb41ceeaa05150d5b0d41f3c66a023 0 299 1363 2024-05-29T23:05:16Z Ben 2 Created page with "Lecture Circuit {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | | | |- | 2024.06.06 | | | |- | 2024.06.05 | | | |- | 2024.06.04 | | | |- | 2024.06..." wikitext text/x-wiki Lecture Circuit {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | | | |- | 2024.06.06 | | | |- | 2024.06.05 | | | |- | 2024.06.04 | | | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} 0e57ffa05ae7fe66e221ccbf8a1c314bb1b3be22 1364 1363 2024-05-29T23:06:14Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | | | |- | 2024.06.05 | | | |- | 2024.06.04 | | | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} e00c98e54afe0b4791963b872c7f3fae3a4a2950 1365 1364 2024-05-29T23:07:44Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | | | |- | 2024.06.05 | | | |- | 2024.06.04 | | | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] 56d1cd595b186705626f2e50768a846a7c77ff60 1367 1365 2024-05-29T23:17:44Z Qdot.me 32 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | | | |- | 2024.06.04 | | | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] ae0fc2cedc89861af9ed407dbdd0341add56cdd7 1368 1367 2024-05-29T23:25:30Z 136.62.52.52 0 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | Hugo | Gaussian Splats | |- | 2024.06.04 | | | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] 03482e9b2a91b342429121a1a8fa6972d8be86b2 1369 1368 2024-05-31T03:30:17Z Budzianowski 19 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | Hugo | Gaussian Splats | |- | 2024.06.04 | Paweł | What I want to believe in | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] 2c4e306f539fd6be13881762eebe857f33003283 1370 1369 2024-05-31T03:56:48Z Budzianowski 19 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | Hugo | Gaussian Splats | |- | 2024.06.04 | Paweł | What I (want to) believe in | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] a1b587162ae840351b83c57cfbbcad8a2e2ffe68 1399 1370 2024-05-31T18:59:59Z Budzianowski 19 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | Add next one | | |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | Hugo | Gaussian Splats | |- | 2024.06.04 | Paweł | What I (want to) believe in | |- | 2024.06.03 | | | |- | 2024.05.31 | Dennis | Speech representation learning papers | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] da37183fd44ae5809f8077010501d271e0dad40f 2 300 1366 2024-05-29T23:16:37Z Qdot.me 32 Created page with "Tom Mloduchowski - Firmware @ KScale [http://qdot.me qdot.me]" wikitext text/x-wiki Tom Mloduchowski - Firmware @ KScale [http://qdot.me qdot.me] 24d97c8b23275b53770049cccf4ae44550b5ed93 0 301 1375 2024-05-31T05:40:56Z Vrtnis 21 Created page with "== Research and Education == '''Research and Educational Applications''' === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics,..." wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. 7a5585a3f15d5f74c07d41da7bce6b1ee82cca99 1376 1375 2024-05-31T05:42:36Z Vrtnis 21 wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. 1e6c41a63011cf3eda78b3f4620277394ab5d4b3 1377 1376 2024-05-31T05:44:06Z Vrtnis 21 wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. ---- == Consumer Applications == '''Home and Personal Assistance Applications''' === Household Chores === * '''Task Performance:''' Assists with household tasks such as picking up items, organizing spaces, and operating household appliances. * '''Customizable Routines:''' Programmed to perform specific routines tailored to individual needs, increasing convenience and efficiency. b8aa6a51150b3989af2fab4ee95d45486ee29c00 1378 1377 2024-05-31T05:44:32Z Vrtnis 21 wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. ---- == Consumer Applications == '''Home and Personal Assistance Applications''' === Household Chores === * '''Task Performance:''' Assists with household tasks such as picking up items, organizing spaces, and operating household appliances. * '''Customizable Routines:''' Programmed to perform specific routines tailored to individual needs, increasing convenience and efficiency. === Assistance for Individuals === * '''Enhanced Independence:''' Provides physical assistance, enabling individuals with mobility impairments to perform daily tasks. * '''Remote Monitoring and Control:''' Allows caregivers to remotely monitor and control the robot, ensuring timely assistance and support. 4a0f8031b3031eecdb48094d329752badd0c0790 1379 1378 2024-05-31T05:44:56Z Vrtnis 21 /* Consumer Applications */ wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. ---- == Consumer Applications == '''Home and Personal Assistance Applications''' === Household Chores === * '''Task Performance:''' Assists with household tasks such as picking up items, organizing spaces, and operating household appliances. * '''Customizable Routines:''' Programmed to perform specific routines tailored to individual needs, increasing convenience and efficiency. === Assistance for Individuals === * '''Enhanced Independence:''' Provides physical assistance, enabling individuals with mobility impairments to perform daily tasks. * '''Remote Monitoring and Control:''' Allows caregivers to remotely monitor and control the robot, ensuring timely assistance and support. === Quality of Life Improvement === * '''Companionship and Interaction:''' Offers social interaction and companionship, improving mental well-being. * '''Safety and Security:''' Monitors home environments for safety hazards and can alert caregivers or emergency services if needed. * '''Adaptive Learning:''' Learns and adapts to user preferences and routines, providing personalized assistance over time. 59abbe009dc6bb1d33f6ad5e1dc2d47a7243ee59 1380 1379 2024-05-31T05:45:04Z Vrtnis 21 /* Consumer Applications */ wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. ---- == Consumer Applications == '''Home and Personal Assistance Applications''' === Household Chores === * '''Task Performance:''' Assists with household tasks such as picking up items, organizing spaces, and operating household appliances. * '''Customizable Routines:''' Programmed to perform specific routines tailored to individual needs, increasing convenience and efficiency. === Assistance for Individuals === * '''Enhanced Independence:''' Provides physical assistance, enabling individuals with mobility impairments to perform daily tasks. * '''Remote Monitoring and Control:''' Allows caregivers to remotely monitor and control the robot, ensuring timely assistance and support. === Quality of Life Improvement === * '''Companionship and Interaction:''' Offers social interaction and companionship, improving mental well-being. * '''Safety and Security:''' Monitors home environments for safety hazards and can alert caregivers or emergency services if needed. * '''Adaptive Learning:''' Learns and adapts to user preferences and routines, providing personalized assistance over time. 7c47e439cc91bdcf34af3a8896d6c60e78556d37 1381 1380 2024-05-31T05:46:40Z Vrtnis 21 wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. ---- == Consumer Applications == '''Home and Personal Assistance Applications''' === Household Chores === * '''Task Performance:''' Assists with household tasks such as picking up items, organizing spaces, and operating household appliances. * '''Customizable Routines:''' Programmed to perform specific routines tailored to individual needs, increasing convenience and efficiency. === Assistance for Individuals === * '''Enhanced Independence:''' Provides physical assistance, enabling individuals with mobility impairments to perform daily tasks. * '''Remote Monitoring and Control:''' Allows caregivers to remotely monitor and control the robot, ensuring timely assistance and support. === Quality of Life Improvement === * '''Companionship and Interaction:''' Offers social interaction and companionship, improving mental well-being. * '''Safety and Security:''' Monitors home environments for safety hazards and can alert caregivers or emergency services if needed. * '''Adaptive Learning:''' Learns and adapts to user preferences and routines, providing personalized assistance over time. == Industrial Applications == '''Industrial Applications''' === Warehouse Logistics === * '''Task Automation:''' Automates repetitive tasks such as picking, packing, and sorting items. * '''Efficient Navigation:''' Capable of navigating through warehouse aisles, reducing time and increasing accuracy in order fulfillment. * '''Real-Time Inventory Management:''' Scans and updates inventory data, ensuring up-to-date stock levels and reducing discrepancies. === Material Handling === * '''Versatile End Effectors:''' Equipped with clamp end effectors for handling medium-sized boxes and containers. * '''Safe Handling:''' Reduces the risk of damage to items by ensuring precise and careful movement. * '''Enhanced Safety:''' Reduces the need for human workers to perform physically demanding and potentially hazardous tasks. === Benefits and Impact === * '''Increased Productivity:''' Faster and more accurate task completion boosts overall warehouse efficiency. * '''Cost Savings:''' Automation reduces labor costs and minimizes errors. * '''Scalability:''' Easily adaptable to different tasks and warehouse configurations. fab916c8c5615ba1b893ad655b562c46c376246d 1382 1381 2024-05-31T05:47:07Z Vrtnis 21 wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. ---- == Consumer Applications == '''Home and Personal Assistance Applications''' === Household Chores === * '''Task Performance:''' Assists with household tasks such as picking up items, organizing spaces, and operating household appliances. * '''Customizable Routines:''' Programmed to perform specific routines tailored to individual needs, increasing convenience and efficiency. === Assistance for Individuals === * '''Enhanced Independence:''' Provides physical assistance, enabling individuals with mobility impairments to perform daily tasks. * '''Remote Monitoring and Control:''' Allows caregivers to remotely monitor and control the robot, ensuring timely assistance and support. === Quality of Life Improvement === * '''Companionship and Interaction:''' Offers social interaction and companionship, improving mental well-being. * '''Safety and Security:''' Monitors home environments for safety hazards and can alert caregivers or emergency services if needed. * '''Adaptive Learning:''' Learns and adapts to user preferences and routines, providing personalized assistance over time. ---- == Industrial Applications == '''Industrial Applications''' === Warehouse Logistics === * '''Task Automation:''' Automates repetitive tasks such as picking, packing, and sorting items. * '''Efficient Navigation:''' Capable of navigating through warehouse aisles, reducing time and increasing accuracy in order fulfillment. * '''Real-Time Inventory Management:''' Scans and updates inventory data, ensuring up-to-date stock levels and reducing discrepancies. === Material Handling === * '''Versatile End Effectors:''' Equipped with clamp end effectors for handling medium-sized boxes and containers. * '''Safe Handling:''' Reduces the risk of damage to items by ensuring precise and careful movement. * '''Enhanced Safety:''' Reduces the need for human workers to perform physically demanding and potentially hazardous tasks. === Benefits and Impact === * '''Increased Productivity:''' Faster and more accurate task completion boosts overall warehouse efficiency. * '''Cost Savings:''' Automation reduces labor costs and minimizes errors. * '''Scalability:''' Easily adaptable to different tasks and warehouse configurations. 6617af73fef27d5826b9e2a18bed158357fe5206 1383 1382 2024-05-31T05:47:27Z Vrtnis 21 wikitext text/x-wiki == Research and Education == '''Research and Educational Applications''' === Research Applications === * '''Human-Robot Interaction Studies:''' Used to study interaction patterns, improving user experience and safety in collaborative environments. * '''Algorithm Development:''' Serves as a testbed for developing and refining robotics algorithms, including motion planning and manipulation. * '''AI and Machine Learning:''' Facilitates research in AI-driven robotics, enabling experiments with machine learning models for autonomous behaviors. === Potential Projects and Experiments === * '''Robotic Path Planning:''' Developing and testing pathfinding algorithms in dynamic environments. * '''Manipulation Tasks:''' Experimenting with different grasping and manipulation techniques using the robot’s end effectors. * '''Sensor Integration:''' Integrating and testing various sensors to enhance robot perception and decision-making. === Educational Tool === * '''STEM Education:''' Provides hands-on learning experiences in robotics, programming, and engineering. * '''Interactive Learning:''' Engages students in interactive projects, enhancing understanding of robotic systems and control algorithms. ---- == Consumer Applications == '''Home and Personal Assistance Applications''' === Household Chores === * '''Task Performance:''' Assists with household tasks such as picking up items, organizing spaces, and operating household appliances. * '''Customizable Routines:''' Programmed to perform specific routines tailored to individual needs, increasing convenience and efficiency. === Assistance for Individuals === * '''Enhanced Independence:''' Provides physical assistance, enabling individuals with mobility impairments to perform daily tasks. * '''Remote Monitoring and Control:''' Allows caregivers to remotely monitor and control the robot, ensuring timely assistance and support. === Quality of Life Improvement === * '''Companionship and Interaction:''' Offers social interaction and companionship, improving mental well-being. * '''Safety and Security:''' Monitors home environments for safety hazards and can alert caregivers or emergency services if needed. * '''Adaptive Learning:''' Learns and adapts to user preferences and routines, providing personalized assistance over time. ---- == Industrial Applications == '''Industrial Applications''' === Warehouse Logistics === * '''Task Automation:''' Automates repetitive tasks such as picking, packing, and sorting items. * '''Efficient Navigation:''' Capable of navigating through warehouse aisles, reducing time and increasing accuracy in order fulfillment. * '''Real-Time Inventory Management:''' Scans and updates inventory data, ensuring up-to-date stock levels and reducing discrepancies. === Material Handling === * '''Versatile End Effectors:''' Equipped with clamp end effectors for handling medium-sized boxes and containers. * '''Safe Handling:''' Reduces the risk of damage to items by ensuring precise and careful movement. * '''Enhanced Safety:''' Reduces the need for human workers to perform physically demanding and potentially hazardous tasks. === Benefits and Impact === * '''Increased Productivity:''' Faster and more accurate task completion boosts overall warehouse efficiency. * '''Cost Savings:''' Automation reduces labor costs and minimizes errors. * '''Scalability:''' Easily adaptable to different tasks and warehouse configurations. f56e24ab4a1b05b5eab62e79d6c658ef4f9d8cfa 0 302 1384 2024-05-31T06:03:45Z Vrtnis 21 Created page with "= Robot Web Viewer = == Overview == The **Robot Web Viewer** provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. Th..." wikitext text/x-wiki = Robot Web Viewer = == Overview == The **Robot Web Viewer** provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using **Reactjs**, **three.js**, and **urdf-loader**. == Features == * **Web-based URDF Viewer**: Allows users to load and visualize URDF files directly in their web browser. * **Simplified Meshes**: Supports simplified meshes for efficient rendering. * **Interactive Controls**: Users can interact with the 3D models using standard web-based controls. * **Easy Setup and Deployment**: The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === ```bash $ git clone https://github.com/vrtnis/robot-web-viewer.git 68a6ce0344b102503abf700ff04ab90b35dbbd4b 1385 1384 2024-05-31T06:03:54Z Vrtnis 21 wikitext text/x-wiki The **Robot Web Viewer** provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using **Reactjs**, **three.js**, and **urdf-loader**. == Features == * **Web-based URDF Viewer**: Allows users to load and visualize URDF files directly in their web browser. * **Simplified Meshes**: Supports simplified meshes for efficient rendering. * **Interactive Controls**: Users can interact with the 3D models using standard web-based controls. * **Easy Setup and Deployment**: The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === ```bash $ git clone https://github.com/vrtnis/robot-web-viewer.git eea93541ac9cfad9a73a783ec4903a233e4a0a38 1386 1385 2024-05-31T06:04:18Z Vrtnis 21 wikitext text/x-wiki The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using **Reactjs**, **three.js**, and **urdf-loader**. == Features == * **Web-based URDF Viewer**: Allows users to load and visualize URDF files directly in their web browser. * **Simplified Meshes**: Supports simplified meshes for efficient rendering. * **Interactive Controls**: Users can interact with the 3D models using standard web-based controls. * **Easy Setup and Deployment**: The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === ```bash $ git clone https://github.com/vrtnis/robot-web-viewer.git a9c647992c0458729d02136ebbd83ff1fc3a11dd 1387 1386 2024-05-31T06:05:25Z Vrtnis 21 wikitext text/x-wiki The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using **Reactjs**, **three.js**, and **urdf-loader**. == Features == * '''Web-based URDF Viewer''': Allows users to load and visualize URDF files directly in their web browser. * '''Simplified Meshes''': Supports simplified meshes for efficient rendering. * '''Interactive Controls''': Users can interact with the 3D models using standard web-based controls. * '''Easy Setup and Deployment''': The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === ```bash $ git clone https://github.com/vrtnis/robot-web-viewer.git 7b639661c75de418b87eb172b99e311a47f05c1d 1388 1387 2024-05-31T06:05:58Z Vrtnis 21 wikitext text/x-wiki The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using '''Reactjs''', '''three.js''', and '''urdf-loader'''. == Features == * '''Web-based URDF Viewer''': Allows users to load and visualize URDF files directly in their web browser. * '''Simplified Meshes''': Supports simplified meshes for efficient rendering. * '''Interactive Controls''': Users can interact with the 3D models using standard web-based controls. * '''Easy Setup and Deployment''': The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === ```bash $ git clone https://github.com/vrtnis/robot-web-viewer.git aa9efb1c8bd72382d84fcc99390156522c4f8698 1389 1388 2024-05-31T06:07:06Z Vrtnis 21 /* Clone the Repository */ wikitext text/x-wiki The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using '''Reactjs''', '''three.js''', and '''urdf-loader'''. == Features == * '''Web-based URDF Viewer''': Allows users to load and visualize URDF files directly in their web browser. * '''Simplified Meshes''': Supports simplified meshes for efficient rendering. * '''Interactive Controls''': Users can interact with the 3D models using standard web-based controls. * '''Easy Setup and Deployment''': The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git </syntaxhighlight> 8fc43c277dc68a214e2958501eb379854fc305df 1391 1389 2024-05-31T06:12:58Z Vrtnis 21 wikitext text/x-wiki [[File:Stompy web viewer.gif|frame|200px|left]] The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using '''Reactjs''', '''three.js''', and '''urdf-loader'''. == Features == * '''Web-based URDF Viewer''': Allows users to load and visualize URDF files directly in their web browser. * '''Simplified Meshes''': Supports simplified meshes for efficient rendering. * '''Interactive Controls''': Users can interact with the 3D models using standard web-based controls. * '''Easy Setup and Deployment''': The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git </syntaxhighlight> bac3695c43d889a8addd72e9715999fe6517ef83 1392 1391 2024-05-31T06:14:59Z Vrtnis 21 wikitext text/x-wiki [[File:Stompy web viewer.gif|frame|200px|left]] The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using '''Reactjs''', '''three.js''', and '''urdf-loader'''. == Features == * '''Web-based URDF Viewer''': Allows users to load and visualize URDF files directly in their web browser. * '''Simplified Meshes''': Supports simplified meshes for efficient rendering. * '''Interactive Controls''': Users can interact with the 3D models using standard web-based controls. * '''Easy Setup and Deployment''': The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git </syntaxhighlight> === Setup === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git $ cd robot-web-viewer/ $ yarn install </syntaxhighlight> ff1a4b8fa15f51c63b53362ea0e9d78c213f54d2 1393 1392 2024-05-31T06:17:01Z Vrtnis 21 wikitext text/x-wiki [[File:Stompy web viewer.gif|frame|200px|left]] The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using '''Reactjs''', '''three.js''', and '''urdf-loader'''. == Features == * '''Web-based URDF Viewer''': Allows users to load and visualize URDF files directly in their web browser. * '''Simplified Meshes''': Supports simplified meshes for efficient rendering. * '''Interactive Controls''': Users can interact with the 3D models using standard web-based controls. * '''Easy Setup and Deployment''': The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git </syntaxhighlight> === Setup === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git $ cd robot-web-viewer/ $ yarn install </syntaxhighlight> === Run === <syntaxhighlight lang="bash"> $ yarn start </syntaxhighlight> 0b0f8ea0eaf1fd05c0c0aa6126a723f38603a7ee 1394 1393 2024-05-31T06:19:21Z Vrtnis 21 wikitext text/x-wiki [[File:Stompy web viewer.gif|frame|200px|left]] The '''Robot Web Viewer''' provides a web-based interface for viewing and interacting with URDF (Unified Robot Description Format) files. The viewer is built using '''Reactjs''', '''three.js''', and '''urdf-loader''' and has been forked and completely updated for use with the latest stable dependencies. [https://github.com/vrtnis/robot-web-viewer https://github.com/vrtnis/robot-web-viewer] == Features == * '''Web-based URDF Viewer''': Allows users to load and visualize URDF files directly in their web browser. * '''Simplified Meshes''': Supports simplified meshes for efficient rendering. * '''Interactive Controls''': Users can interact with the 3D models using standard web-based controls. * '''Easy Setup and Deployment''': The project includes scripts for easy local setup and deployment. == Setup Instructions == To set up the Robot Web Viewer locally, follow these steps: === Clone the Repository === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git </syntaxhighlight> === Setup === <syntaxhighlight lang="bash"> $ git clone https://github.com/vrtnis/robot-web-viewer.git $ cd robot-web-viewer/ $ yarn install </syntaxhighlight> === Run === <syntaxhighlight lang="bash"> $ yarn start </syntaxhighlight> d4e4ab6fb1826ccbddd11404670ef048524353d0 6 303 1390 2024-05-31T06:10:51Z Vrtnis 21 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 1 1395 1355 2024-05-31T16:28:35Z Budzianowski 19 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Tesla]] | [[Optimus]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[System Technology Works]] | [[ZEUS2Q]] | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Physical Intelligence]] | |- | [[Skild]] | |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[WorkFar]] | [[WorkFar Syntro]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 94bc0861059a9da6746d6fb2d4ec8a0dde1b50e6 1396 1395 2024-05-31T16:30:40Z Budzianowski 19 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Tesla]] | [[Optimus]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[System Technology Works]] | [[ZEUS2Q]] | [[K-Scale Labs]] | [[Stompy]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Kepler]] | [[K1]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Kayra.org]] | [[Kayra]] |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[WorkFar]] | [[WorkFar Syntro]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} dc0713097deef34fe07ec58b19eecbf397abd309 1397 1396 2024-05-31T16:33:27Z Budzianowski 19 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Humanoid Robots === {| class="wikitable" |- ! Company ! Robots |- | [[Anthrobotics]] | [[Anthro]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[DREAME]] | [[MagicBot]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Xpeng]] | [[PX5]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[Proxy]] | |- | [[SUPCON]] | [[Navigator α]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[WorkFar]] | [[WorkFar Syntro]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} ee4a335a41b0db2b589bbba475aa59955a96d22e 1398 1397 2024-05-31T16:35:09Z Budzianowski 19 /* List of Humanoid Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 830cbdbe875817b090703b80231c1e3d5e908860 0 294 1400 1346 2024-05-31T19:02:17Z 108.211.178.220 0 wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://twitter.com/kscalelabs/status/1788968705378181145 2024.05.10] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |} [[Category:K-Scale]] 1c7682457870cd831a81feb76d073e0af268e03f 0 2 1401 1071 2024-06-01T04:25:54Z Dymaxion 22 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] * [[Gripper History]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions === Conventions === The images below show our pin convention for the CAN bus when using various connectors. <gallery> Kscale db9 can bus convention.jpg Kscale phoenix can bus convention.jpg </gallery> = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. = Artwork = Here's some art of Stompy! <gallery> Stompy 1.png Stompy 2.png Stompy 3.png Stompy 4.png </gallery> [[Category:Robots]] [[Category:Open Source]] [[Category:K-Scale]] ff7854d1e249f1aa6d72cfaa47f51d43aed4b25d 0 304 1402 2024-06-01T04:26:16Z Dymaxion 22 Created page with "==Gripper Design Iterations== Original Design: [[File:Gripper 0.jpg|thumb|The non-optimized gripper based on the UMI study]] Problems: Too flexible at fingertips, too stiff in..." wikitext text/x-wiki ==Gripper Design Iterations== Original Design: [[File:Gripper 0.jpg|thumb|The non-optimized gripper based on the UMI study]] Problems: Too flexible at fingertips, too stiff in inner surface Updated version 1: [[File:Gripper version 1.jpg|thumb|Updated version 1]] Made ribs thinner at their ends so that they would bend more easily Added triangular cuts along the inner surface to allow the material to flex Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear Updated version 2: [[File:Stompy gripper 2.jpg|thumb|Updated Version 2]] Moved slots to only align with the three thickest ribs, reducing potential for tears Created an angled profile where the tips of the grippers are narrower than the base Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases Updated version 3: [[File:Stompy gripper version 3.jpg|thumb|Updated version 3]] Decreased hole size to allow for better threading Problems: tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail) Updated version 4: [[File:Gripper version 4.jpg|thumb|Updated version 4]] Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more Problems: too flimsy in general, and particularly at tips Updated version 5: [[File:Updated version 5.jpg|thumb|Updated version 5]] Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports Problems: poor print quality due to lack of supports Updated version 6: Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward Updated version 7: Decreased profile of rib material to make screw installation process easier 34b68c7d70de79faae6f2aea040e6375977afc43 6 305 1403 2024-06-01T04:28:22Z Dymaxion 22 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 306 1404 2024-06-01T04:30:14Z Dymaxion 22 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 307 1405 2024-06-01T04:32:24Z Dymaxion 22 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 308 1406 2024-06-01T04:34:23Z Dymaxion 22 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 309 1407 2024-06-01T04:35:01Z Dymaxion 22 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 310 1408 2024-06-01T04:35:37Z Dymaxion 22 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 304 1409 1402 2024-06-01T04:42:37Z Dymaxion 22 /* Gripper Design Iterations */ wikitext text/x-wiki ==Gripper Design Iterations== Original Design: [[File:Gripper 0.jpg|thumb|The non-optimized gripper based on the UMI study]] Problems: Too flexible at fingertips, too stiff in inner surface Updated version 1: [[File:Gripper version 1.jpg|thumb|Updated version 1]] Made ribs thinner at their ends so that they would bend more easily Added triangular cuts along the inner surface to allow the material to flex Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear Updated version 2: [[File:Stompy gripper 2.jpg|thumb|Updated Version 2]] Moved slots to only align with the three thickest ribs, reducing potential for tears Created an angled profile where the tips of the grippers are narrower than the base Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases Updated version 3: [[File:Stompy gripper version 3.jpg|thumb|Updated version 3]] Decreased hole size to allow for better threading Problems: tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail) Updated version 4: [[File:Gripper version 4.jpg|thumb|Updated version 4]] Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more Problems: too flimsy in general, and particularly at tips Updated version 5: [[File:Updated version 5.jpg|thumb|Updated version 5]] Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports Problems: poor print quality due to lack of supports Updated version 6: Used solid tip and fewer ribs in the other section. Problems: Tip will not bend around objects Assembly with screws remained awkward Updated version 7: Decreased profile of rib material to make screw installation process easier 30e02f43adb0c63a56cc0399234bf3a0ea5d515e 1410 1409 2024-06-01T05:03:39Z Dymaxion 22 wikitext text/x-wiki == Gripper Design Iterations == === Original Design === Problems: Too flexible at fingertips, too stiff in inner surface. [[File:Gripper 0.jpg|none|300px|The non-optimized gripper based on the UMI study]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Decreased profile of rib material to make screw installation process easier. 3216e11c9ce591b54fe581111879d7bb472de824 1413 1410 2024-06-02T06:48:42Z Dymaxion 22 /* Updated Version 7 */ wikitext text/x-wiki == Gripper Design Iterations == === Original Design === Problems: Too flexible at fingertips, too stiff in inner surface. [[File:Gripper 0.jpg|none|300px|The non-optimized gripper based on the UMI study]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 ===Switched cuts to the inside wall and eliminated redundant rib. 53b89cf0de524897b257f20dceae497eae52156c 1414 1413 2024-06-02T06:52:21Z Dymaxion 22 /* Updated Version 6 */ wikitext text/x-wiki == Gripper Design Iterations == === Original Design === Problems: Too flexible at fingertips, too stiff in inner surface. [[File:Gripper 0.jpg|none|300px|The non-optimized gripper based on the UMI study]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. 6b85cab78ddf68f2f41a92667bb29fbaa4edd5b3 1415 1414 2024-06-02T08:15:19Z Dymaxion 22 /* Original Design */ wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. 43d1824872c0f4e7f199599254554ff5ffb3fcac 1417 1415 2024-06-02T08:17:03Z Dymaxion 22 /* Updated Version 7 */ wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper 7.jpg|none|300px|Updated version 7]] 9acc0789755e8a791dd0d7b55fbd3d8d2bdc3625 1419 1417 2024-06-02T08:21:08Z Dymaxion 22 /* Updated Version 7 */ wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper 7.jpg|thumb]] aae55da46fad7e56cea38f6f69fc7e5ee585ff0e 1420 1419 2024-06-02T08:28:27Z Dymaxion 22 /* Updated Version 7 */ wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] 6799f26f51363f7989fa52d8b602669a5f624b76 1421 1420 2024-06-02T08:29:10Z Dymaxion 22 /* Updated Version 7 */ wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper 7.jpg|none|300px|Updated version 7]] 9acc0789755e8a791dd0d7b55fbd3d8d2bdc3625 1422 1421 2024-06-02T08:29:29Z Dymaxion 22 /* Updated Version 7 */ wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stomy gripper 7.jpg|none|300px|Updated version 7]] c219b615236204b6f4eeb05ba31755c48c96bc28 1428 1422 2024-06-03T21:48:48Z Dymaxion 22 wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stompy gripper version 7.jpg|none|300px|Updated version 7]] === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stomy gripper 7.jpg|none|300px|Updated version 7]] === Updated Version 8 === Tapered inside wall and extended it to the now restored rib. [[File:Stompy gripper 8.jpg|thumb|Updated version 8]] 99e04c142a5d1af0fbdc6141461c8deeec06a7b9 1429 1428 2024-06-03T21:49:35Z Dymaxion 22 wikitext text/x-wiki == Gripper Design Iterations == === Updated Version 1 === Made ribs thinner at their ends so that they would bend more easily. Added triangular cuts along the inner surface to allow the material to flex. Problems: The triangular cut closest to the tip caused the inner surface of the grippers to tear. [[File:Gripper version 1.jpg|none|300px|Updated version 1]] === Updated Version 2 === Moved slots to only align with the three thickest ribs, reducing potential for tears. Created an angled profile where the tips of the grippers are narrower than the base. Problems: Screw holes were too large for screws to be inserted from below, tears persisted in some cases. [[File:Stompy gripper 2.jpg|none|300px|Updated Version 2]] === Updated Version 3 === Decreased hole size to allow for better threading. Problems: Tears persisted, insufficient clearance for hex key to actually tighten the screws from above (necessary for use with slide rail). [[File:Stompy gripper version 3.jpg|none|300px|Updated version 3]] === Updated Version 4 === Decreased profile of rib material to make screw installation process easier, decreased cross section of individual ribs to allow them to bend more. Problems: Too flimsy in general, and particularly at tips. [[File:Gripper version 4.jpg|none|300px|Updated version 4]] === Updated Version 5 === Restored rib material and eliminated one triangular cut to solve tearing issue, attempted to print entirely without supports. Problems: Poor print quality due to lack of supports. [[File:Updated version 5.jpg|none|300px|Updated version 5]] === Updated Version 6 === Used solid tip and fewer ribs in the other section. Assembly with screws remained awkward. [[File:Stompy gripper version 6.jpg|none|300px|Updated version 6]] === Updated Version 7 === Switched cuts to the inside wall and eliminated redundant rib. [[File:Stomy gripper 7.jpg|none|300px|Updated version 7]] === Updated Version 8 === Tapered inside wall and extended it to the now restored rib. [[File:Stompy gripper 8.jpg|none|300px|Updated version 8]] 24dd4804ccec62a6f3f09a71348277ab1aa52126 6 311 1411 2024-06-01T05:04:55Z Dymaxion 22 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 1 1412 1398 2024-06-01T22:06:50Z 108.211.178.220 0 /* Getting Started */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 67468244a9e6a8cefbef04920b9b05dd94ba3e47 1453 1412 2024-06-06T00:28:49Z Admin 1 /* List of Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 8ac879d0730ee9a580fdbf87a1b857d995864f0b 6 312 1416 2024-06-02T08:16:38Z Dymaxion 22 wikitext text/x-wiki A photo of one stompy gripper with a large lower void. 529dda8186713569c55d73c3ed8cc844d0629c90 6 313 1418 2024-06-02T08:18:21Z Dymaxion 22 A photo of one stompy gripper with a large void and triangles on the back side of the gripping face. wikitext text/x-wiki == Summary == A photo of one stompy gripper with a large void and triangles on the back side of the gripping face. 21d70f088195ac880a2af9b0b835cc9b3d6caca0 0 299 1423 1399 2024-06-03T17:35:55Z 108.211.178.220 0 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | Add next one | | |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | Hugo | Gaussian Splats | |- | 2024.06.04 | Paweł | What I (want to) believe in | |- | 2024.06.03 | Dennis | Speech representation learning papers | |- | 2024.05.31 | | | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] b903b9676410d1c64ae441eb634afea133480d0a 1430 1423 2024-06-03T22:21:55Z Budzianowski 19 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | Add next one | | |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | Speech representation learning papers | |- | 2024.06.03 | Paweł | What I (want to) believe in | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] 07e0d9775a820808805c546b4f6f20789443d72a 1444 1430 2024-06-05T01:52:42Z Admin 1 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic ! Link |- | | Timothy | NeRFs | |- | 2024.06.07 | Ben | Quantization | |- | 2024.06.06 | Tom | Linux Raw | |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | Speech representation learning papers | |- | 2024.06.03 | Paweł | What I (want to) believe in | |- | 2024.05.30 | Isaac | VLMs | |- | 2024.05.29 | Allen | PPO | |} [[Category: K-Scale]] 6e56a3ae939c10ebf1f48ec2fa5f36c801023309 1445 1444 2024-06-05T01:53:03Z Admin 1 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | | Timothy | NeRFs |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | Speech representation learning papers |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 5266ddb69aca91df94ba40c603841667b02545a5 0 281 1424 1374 2024-06-03T18:26:08Z Vrtnis 21 /*Add BHR-4*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ - | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 28b6c494ffcb6d7810ec0cbd0b49386bf49e78d6 1425 1424 2024-06-03T18:27:08Z Vrtnis 21 /*Add BHR-4*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 07deea957a246684a127cdfc9ad7653dc51ad538 1426 1425 2024-06-03T18:32:04Z Vrtnis 21 /*Add Tiago and InMoov*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology b1e4a5cc3c0f88a0e31cd7b63a7dcef996cf7b35 1446 1426 2024-06-05T19:37:41Z Vrtnis 21 /*Add OpenAI BipedalWalker*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology 00763d39c5ba867005ad46f641bb827a79bddd63 1447 1446 2024-06-05T19:42:07Z Vrtnis 21 /* References */ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology a6d542853f67af75cfc04c9966cfc29c5b9b4fd7 1448 1447 2024-06-05T19:47:08Z Vrtnis 21 /*Add cite*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } b68b27443341884a635a6e2be3c96ef3a0fc3348 1449 1448 2024-06-05T19:47:42Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == {@misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } d37ec6e202b14ba5f6c39655a57fb30f1c01827f 1450 1449 2024-06-05T19:50:53Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Authors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> 18b54daea8139fbc158420f7084d3d022dd8e57d 1451 1450 2024-06-05T19:51:08Z Vrtnis 21 /* Citation */ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> 778cae7cbe49415b938a21116b7d71a8baae90f2 1452 1451 2024-06-05T19:53:07Z Vrtnis 21 /*Add BOLT*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> 446c7eb347cf6d1486ce9b69c1714116a2d50371 6 314 1427 2024-06-03T21:47:58Z Dymaxion 22 wikitext text/x-wiki Two stompy grippers screwed to another part. 01f93ee70c12597b5f354b4ffc1d69448883c442 0 315 1431 2024-06-04T00:33:26Z Vedant 24 Created page with "= Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB...." wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev andthe board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the e77d9c0f90d864de947fc8cd7b5336f43a483839 1432 1431 2024-06-04T00:34:33Z 108.211.178.220 0 wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev andthe board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the linux terminal itself. 7ab0016f09aa045bc7240b59fd1d7ffd9cceaa01 1433 1432 2024-06-04T00:35:52Z 108.211.178.220 0 wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev andthe board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the linux terminal itself. 43957c0152a5c1707928287f61556ae75b891f46 1434 1433 2024-06-04T00:36:07Z 108.211.178.220 0 /* For the Jetson Orin Nano */ wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev andthe board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the linux terminal itself. 7c3e0d5c465aa1b20aba6d77c3b39a981a7034d5 1435 1434 2024-06-04T00:38:01Z 108.211.178.220 0 /* Flashing */ wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev and the board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the linux terminal itself. Current Game Plan: - mess around with the programming of the GPIO pins: Figure out if there are ways to choose access teh data that the GPIO pins are Or Receiving. Test if it is possible to reconfigure the pins on the jetson on the firmware side db5cd86b096e03b7fd3f16425217135076e291d9 1438 1435 2024-06-04T09:47:13Z 108.211.178.220 0 wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev and the board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the linux terminal itself. Current Game Plan: - mess around with the programming of the GPIO pins: Figure out if there are ways to choose access teh data that the GPIO pins are Or Receiving. Test if it is possible to reconfigure the pins on the jetson on the firmware side Build Time: - On a single Nvidia Nano, it takes about 45 mins - 1 hour to complete the build. The build is encrypted through RSA. The source is still accesssible and every time changes are made to the source files and want ot be reflected, the build files have to be remade. Current goal (Figure out if there's a way to make specific build files to decrease development time). Notes: - Need to install libssl-dev (depends on whether the certain packages are included when running the script) 32c769652bba1b1457e0df3c11d74abeab858209 1442 1438 2024-06-04T18:23:46Z Goblinrum 25 wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev and the board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the linux terminal itself. Current Game Plan: - mess around with the programming of the GPIO pins: Figure out if there are ways to choose access teh data that the GPIO pins are Or Receiving. Test if it is possible to reconfigure the pins on the jetson on the firmware side Build Time: - On a single Nvidia Nano, it takes about 45 mins - 1 hour to complete the build. The build is encrypted through RSA. The source is still accessible and every time changes are made to the source files and want ot be reflected, the build files have to be remade. Current goal (Figure out if there's a way to make specific build files to decrease development time). Notes: - General requirements to build Linux kernel still apply: e.g. `build-essentials`, `bc`, `libssl-dev`, etc - Need to install `libssl-dev (depends on whether the certain packages are included when running the script) 66661a6a82fb7bfd82722736aaa29b96f343e2c8 0 316 1436 2024-06-04T07:42:40Z Ben 2 Created page with "=== Links === * [https://developer.nvidia.com/embedded/jetson-linux NVIDIA Jetson Linux Driver Package]" wikitext text/x-wiki === Links === * [https://developer.nvidia.com/embedded/jetson-linux NVIDIA Jetson Linux Driver Package] 7d5230b8f59b0bb86246cbba31615219ec96c270 1437 1436 2024-06-04T07:43:07Z Ben 2 wikitext text/x-wiki === Links === * [https://developer.nvidia.com/embedded/jetson-linux NVIDIA Jetson Linux Driver Package] * [https://docs.nvidia.com/jetson/archives/r35.4.1/DeveloperGuide/text/SD/Kernel/KernelCustomization.html Kernel customization] e2006a6f5a6f42c314e3bd2b69e1ef2943faafcd 1443 1437 2024-06-04T18:54:53Z Goblinrum 25 wikitext text/x-wiki === Notes === You need to install the following if you are not compiling and building on the nano itself (standard for any attempt to build the Linux kernel yourself) sudo apt-get install git fakeroot build-essential ncurses-dev xz-utils libssl-dev bc flex libelf-dev bison Also, install the toolchain here: https://docs.nvidia.com/jetson/archives/r35.4.1/DeveloperGuide/text/AT/JetsonLinuxToolchain.html#at-jetsonlinuxtoolchain === Links === * [https://developer.nvidia.com/embedded/jetson-linux NVIDIA Jetson Linux Driver Package] * [https://docs.nvidia.com/jetson/archives/r35.4.1/DeveloperGuide/text/SD/Kernel/KernelCustomization.html Kernel customization] 02a25909d14c2319c038ddd32e5b866b10d72db9 0 218 1439 982 2024-06-04T16:00:54Z Tom 23 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run <code>sudo ./tools/l4t_flash_prerequisites.sh</code> * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> under NVIDIA CORP. APX * Run <code>./flash.sh </code> Different options for different usecases(https://docs.nvidia.com/jetson/archives/r36.3/DeveloperGuide/IN/QuickStart.html#in-quickstart) Jetson AGX Orin Developer Kit (eMMC): $ sudo ./flash.sh jetson-agx-orin-devkit internal * Watch for few minutes, typically it crashes early, then go for lunch. [[Category: Firmware]] 9786dd856928865a0d28510f06ea0918e1e8d9da 1440 1439 2024-06-04T16:01:48Z Tom 23 /* BSP approach (avoids SDK Manager) */ wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run <code>sudo ./tools/l4t_flash_prerequisites.sh</code> * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> under NVIDIA CORP. APX * Run <code>./flash.sh </code> Different options for different usecases(https://docs.nvidia.com/jetson/archives/r36.3/DeveloperGuide/IN/QuickStart.html#in-quickstart) Jetson AGX Orin Developer Kit (eMMC): $ <code>sudo ./flash.sh jetson-agx-orin-devkit internal</code> * Watch for few minutes, typically it crashes early, then go for lunch. [[Category: Firmware]] e6421d3859487babd61235a1ada3dac850a6b28e 1441 1440 2024-06-04T16:02:01Z Tom 23 /* BSP approach (avoids SDK Manager) */ wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run <code>sudo ./tools/l4t_flash_prerequisites.sh</code> * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> under NVIDIA CORP. APX * Run <code>./flash.sh </code> Different options for different usecases(https://docs.nvidia.com/jetson/archives/r36.3/DeveloperGuide/IN/QuickStart.html#in-quickstart) Jetson AGX Orin Developer Kit (eMMC): <code>$ sudo ./flash.sh jetson-agx-orin-devkit internal</code> * Watch for few minutes, typically it crashes early, then go for lunch. [[Category: Firmware]] 05bfbb05718edf35c0cb2a572ce427836def3078 0 317 1454 2024-06-06T00:29:57Z Admin 1 Created page with "[https://mirsee.com/ Pollen Robotics] is a robotics company based in Bordeaux, France. {{infobox company | name = Pollen Robotics | country = France | website_link = https:/..." wikitext text/x-wiki [https://mirsee.com/ Pollen Robotics] is a robotics company based in Bordeaux, France. {{infobox company | name = Pollen Robotics | country = France | website_link = https://www.pollen-robotics.com/ | robots = [[Reachy]] }} [[Category:Companies]] 44c4b394b648451fc812b3b96ce4325ee84e9c97 1455 1454 2024-06-06T00:30:04Z Admin 1 wikitext text/x-wiki [https://mirsee.com/ Pollen Robotics] is a robotics company based in Bordeaux, France. {{infobox company | name = Pollen Robotics | country = France | website_link = https://www.pollen-robotics.com/ | robots = [[Reachy]] }} [[Category:Companies]] 70f45176cc6c80de5b95aea706e85af85078f62a 0 194 1456 811 2024-06-06T00:30:09Z Admin 1 wikitext text/x-wiki [https://mirsee.com/ Mirsee Robotics] is a robotics company based in Cambridge, ON, Canada. Along with custom actuators, hands, and other solutions, they have developed two humanoid robots, [[Beomni]] and [[Mirsee]], both of which have wheeled bases. {{infobox company | name = Mirsee Robotics | country = Canada | website_link = https://mirsee.com/ | robots = [[Beomni]], [[Mirsee]] }} [[Category:Companies]] 8c1dc59aeabda52c05ae904ee9554b4e2695861c 0 318 1457 2024-06-06T00:32:35Z Admin 1 Created page with "Beomni is a humanoid robot developed by [[Mirsee Robotics]] for the Beyond Imagination AI company: https://www.beomni.ai/. {{infobox robot | name = Reachy | organization = ..." wikitext text/x-wiki Beomni is a humanoid robot developed by [[Mirsee Robotics]] for the Beyond Imagination AI company: https://www.beomni.ai/. {{infobox robot | name = Reachy | organization = [[Pollen Robotics]] | video_link = https://www.youtube.com/watch?v=oZxHkp4-DnM }} [[Category:Robots]] 972d8feb8b8445b96e107215be1efbe362a76a84 0 318 1458 1457 2024-06-06T00:32:46Z Admin 1 wikitext text/x-wiki Reachy is a humanoid robot developed by [[Pollen Robotics]]. {{infobox robot | name = Reachy | organization = [[Pollen Robotics]] | video_link = https://www.youtube.com/watch?v=oZxHkp4-DnM }} [[Category:Robots]] c00be9dac547129ddbabf2dc77ee79d92dbfdd7b 0 2 1459 1401 2024-06-06T00:34:59Z Admin 1 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb|Stompy standing up]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] * [[Gripper History]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions === Conventions === The images below show our pin convention for the CAN bus when using various connectors. <gallery> Kscale db9 can bus convention.jpg Kscale phoenix can bus convention.jpg </gallery> = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. = Artwork = Here's some art of Stompy! <gallery> Stompy 1.png Stompy 2.png Stompy 3.png Stompy 4.png </gallery> [[Category:Robots]] [[Category:Open Source]] [[Category:K-Scale]] 9cf6215dee47b745206fbcab51f3fb170035bc49 0 317 1460 1455 2024-06-06T00:35:15Z Admin 1 wikitext text/x-wiki [https://mirsee.com/ Pollen Robotics] is a robotics company based in Bordeaux, France. {{infobox company | name = Pollen Robotics | country = France | website_link = https://www.pollen-robotics.com/ | robots = [[Reachy]] }} They have published many open source designs [https://www.pollen-robotics.com/opensource/ here]. [[Category:Companies]] 308e84f5532bf9dd9f34a7d010e442aa0cfa9064 0 281 1461 1452 2024-06-06T05:55:32Z Vrtnis 21 /* HRP-4 */ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> db0c1f0ffdae6b708f9253cfef6a81e6c0e639c3 1462 1461 2024-06-06T05:56:47Z Vrtnis 21 /* PI-4 */ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 43 || PI4 || PAL Robotics || URDF || [https://github.com/pal-robotics/pi4_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> 1cb6b76b407862bf9604799a66d54900a6a23a5a 1463 1462 2024-06-06T05:58:24Z Vrtnis 21 /* HRP-7P */ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 43 || PI4 || PAL Robotics || URDF || [https://github.com/pal-robotics/pi4_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 44 || HRP-7P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp7p_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> 9c9918612d4204052247c6720a32707b1a4b5fc7 1464 1463 2024-06-06T21:32:00Z Vrtnis 21 /*Add Juno*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 43 || PI4 || PAL Robotics || URDF || [https://github.com/pal-robotics/pi4_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 44 || HRP-7P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp7p_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 45 || Juno || UC Berkeley || URDF || [https://github.com/BerkeleyAutomation/juno_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> 04dc6bb44b8fcacde48079c79e9aac7e008548f6 1465 1464 2024-06-06T21:32:35Z Vrtnis 21 /* Add Leo and Apollo */ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 43 || PI4 || PAL Robotics || URDF || [https://github.com/pal-robotics/pi4_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 44 || HRP-7P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp7p_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 45 || Juno || UC Berkeley || URDF || [https://github.com/BerkeleyAutomation/juno_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 46 || Leo || Georgia Institute of Technology || URDF || [https://github.com/GeorgiaTechRobotLearning/leo_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 47 || Apollo || University of Pennsylvania || URDF || [https://github.com/penn-robotics/apollo_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> cdf19c0dfc2b5d4fa4204b28710fad014349c364 1507 1465 2024-06-11T19:26:20Z Vrtnis 21 /*Scuttle*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 43 || PI4 || PAL Robotics || URDF || [https://github.com/pal-robotics/pi4_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 44 || HRP-7P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp7p_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 45 || Juno || UC Berkeley || URDF || [https://github.com/BerkeleyAutomation/juno_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 46 || Leo || Georgia Institute of Technology || URDF || [https://github.com/GeorgiaTechRobotLearning/leo_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 47 || Apollo || University of Pennsylvania || URDF || [https://github.com/penn-robotics/apollo_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 48 || Scuttle || Open Robotics || URDF || [https://github.com/openrobotics/scuttle_description URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> eac2b245114e17402b638f7016b338aa53e7f799 0 319 1466 2024-06-06T21:34:32Z Vrtnis 21 /*Add Pose Estimation Overview*/ wikitext text/x-wiki '''Pose estimation''' is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. These models can range from simple algorithms for 2D pose estimation to more complex systems that infer 3D poses. Recent advances in deep learning have significantly improved the accuracy and robustness of pose estimation systems, enabling their use in real-time applications. 708c04cfa0086b79906bb1e353806ca6df0aa7eb 1467 1466 2024-06-06T21:37:59Z Vrtnis 21 /*Add Mediapipe*/ wikitext text/x-wiki '''Pose estimation''' is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. These models can range from simple algorithms for 2D pose estimation to more complex systems that infer 3D poses. Recent advances in deep learning have significantly improved the accuracy and robustness of pose estimation systems, enabling their use in real-time applications. '''MediaPipe''' is an advanced computer vision tool developed by Google, designed to accurately identify and track human poses in real-time. MediaPipe leverages machine learning to detect and map out keypoints on the human body, such as the elbows, knees, and shoulders, providing a detailed understanding of body posture and movements. 9be84cec7f9d20e79409756d35cd45bfe9fad59a 1468 1467 2024-06-07T01:49:20Z Vrtnis 21 /*Add OpenPose*/ wikitext text/x-wiki {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |} fa11b6aafe30053b5827ce5368d3da68c47b82eb 1469 1468 2024-06-07T01:50:00Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. These models can range from simple algorithms for 2D pose estimation to more complex systems that infer 3D poses. Recent advances in deep learning have significantly improved the accuracy and robustness of pose estimation systems, enabling their use in real-time applications. {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |} 654ff331da09fddc9b2906ca22a6e26f7d0640a8 1470 1469 2024-06-07T01:50:22Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |} a939b8c4fa2caf38643ba2844e75ac1c5cf79a25 1471 1470 2024-06-07T01:55:55Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |} de9436a67134f4369a58e972bf5a40b15cff09cb 1472 1471 2024-06-07T01:56:31Z Vrtnis 21 /* Add MoveNet */ wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 2 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |} d158e6ff80e610698aa3e688b34a5a621276d78c 1473 1472 2024-06-07T04:51:42Z Vrtnis 21 /*MediaPipe and Detectron2 */ wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [ https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 || ✔️ || ✔️ || ✔️ |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [ https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 || ✔️ || ✔️ || ✔️ |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |} 32f8c5992ea2abb52ad71689a859616f9622782c 1474 1473 2024-06-07T04:52:39Z Vrtnis 21 /* Pose Estimation Related Models */ wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |} 91fccfeca84b99b28b80a2be8dd37f814d726180 1475 1474 2024-06-07T05:01:52Z Vrtnis 21 /* Pose Estimation Related Models */ wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} e6cdecf2ea318650d2baf3648aa2a6cb5279cfe5 1477 1475 2024-06-07T05:38:08Z Vrtnis 21 /*Add mediapipe detection*/ wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} [[File:Pose detection overlay.gif|720px|thumb|Mediapipe Pose Detection]] 6330014352f7b417d0d3787511a7253a23dbb2e9 1479 1477 2024-06-07T05:43:53Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} [[File:Pose detection overlay.gif|720px|thumb|Mediapipe Pose Detection]] [[File:Poseoutput.gif|720px|thumb|Mediapipe Pose Detection]] a4116bb6c11e770ae1357e414fd43cfffda37fdc 1480 1479 2024-06-07T05:44:28Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} [[File:Pose detection overlay.gif|720px|thumb|Mediapipe Pose Detection]] 6330014352f7b417d0d3787511a7253a23dbb2e9 1485 1480 2024-06-07T05:50:50Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} [[File:Pose detection overlay.gif|720px|thumb|Mediapipe Pose Detection]] <gallery> Pose_example1.png|About to Stand Pose_example2.png|Standing but error in leg detection </gallery> edf7a9df84d8340a37d96653323320eabde30a2f 1486 1485 2024-06-07T05:51:56Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} [[File:Pose detection overlay.gif|720px|thumb|Mediapipe Pose Detection]] <gallery> Pose_example1.png|About to Stand Pose_example2.png|Standing but error in leg detection Pose_example3.png|Standing but error in leg detection Pose_example4.png|Standing but error in leg detection </gallery> 71e3aa93c50fbb326f1f40b468f543e7496ee7bd 1487 1486 2024-06-07T05:53:04Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} [[File:Pose detection overlay.gif|720px|thumb|Mediapipe Pose Detection]] <gallery> Pose_example1.png|About to Stand Pose_example2.png|Standing but error in leg detection Pose_example3.png|Foreground missed Pose_example4.png|Hoodie </gallery> af233217b14dc3e4f2e644d7213965757c8e0a5f 1489 1487 2024-06-07T06:47:54Z Vrtnis 21 wikitext text/x-wiki Pose estimation is a computer vision technique that predicts the configuration of a person's or object's joints or parts in an image or video. It involves detecting and tracking the position and orientation of these parts, usually represented as keypoints. Pose estimation is widely used in applications such as motion capture, human-computer interaction, augmented reality, and robotics. The process typically involves training machine learning models on large datasets of annotated images to accurately identify and locate the keypoints. === Pose Estimation Related Models === {| class="wikitable sortable" |- ! Sr No !! Model !! Developer !! Key Points !! Source !! License |- | 1 || MediaPipe || Google || Tracking 33 key points on the human body, offering cross-platform, customizable ML solutions || [https://github.com/google/mediapipe MediaPipe GitHub] || Apache 2.0 |- | 2 || Detectron2 || Facebook AI Research || High-performance codebase for object detection and segmentation, including pose estimation || [https://github.com/facebookresearch/detectron2 Detectron2 GitHub] || Apache 2.0 |- | 3 || OpenPose || Carnegie Mellon University || Detecting key points of the human body, including hand, facial, and foot || [https://github.com/CMU-Perceptual-Computing-Lab/openpose OpenPose GitHub] || MIT |- | 4 || MoveNet || Google Research || Detecting 17 critical key points of the human body || [https://github.com/tensorflow/tfjs-models/tree/master/posenet MoveNet GitHub] || Apache 2.0 |- | 5 || PoseNet || Google Research || Detecting different body parts, providing comprehensive skeletal information || [https://github.com/tensorflow/tfjs-models/tree/master/posenet PoseNet GitHub] || Apache 2.0 |- | 6 || DCPose || Deep Dual Consecutive Network || Detecting human pose from multiple frames, addressing motion blur and occlusions || [https://github.com/DeepDualConsecutivePose/dcpose DCPose GitHub] || MIT |- | 7 || DensePose || Facebook AI Research || Mapping human-based pixels from an RGB image to the 3D surface of a human body || [https://github.com/facebookresearch/DensePose DensePose GitHub] || Apache 2.0 |- | 8 || HigherHRNet || HRNet || Addressing scaling differences in pose prediction, especially for shorter people || [https://github.com/HRNet/HigherHRNet HigherHRNet GitHub] || MIT |- | 9 || Lightweight OpenPose || Daniil-Osokin || Real-time inference with minimal accuracy drop, detecting human poses through key points || [https://github.com/Daniil-Osokin/lightweight-human-pose-estimation Lightweight OpenPose GitHub] || MIT |- | 10 || AlphaPose || MVIG-SJTU || Detecting multiple individuals in various scenes, achieving high mAP on COCO and MPII datasets || [https://github.com/MVIG-SJTU/AlphaPose AlphaPose GitHub] || MIT |} [[File:Pose detection overlay.gif|720px|thumb|Mediapipe Pose Detection]] <gallery> Pose_example1.png|About to Stand Pose_example2.png|Standing but error in leg detection Pose_example3.png|Foreground missed Pose_example4.png|Hoodie </gallery> [[File:Poseoutput white orig.gif|720px|thumb|Mediapipe Pose Detection]] 776a81ffcec351c1022ca92de74dd2ba5130e608 6 320 1476 2024-06-07T05:37:13Z Vrtnis 21 wikitext text/x-wiki Pose detection overlay 37adfb1a1a961e6f65f8d983c6cb774f8a2c6a16 6 321 1478 2024-06-07T05:40:43Z Vrtnis 21 wikitext text/x-wiki Mediapipe pose detection without background 7d07ad192169da0767128c6cb33488f1b06b1c00 6 322 1481 2024-06-07T05:45:49Z Vrtnis 21 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 323 1482 2024-06-07T05:46:23Z Vrtnis 21 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 324 1483 2024-06-07T05:47:17Z Vrtnis 21 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 325 1484 2024-06-07T05:47:48Z Vrtnis 21 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 6 326 1488 2024-06-07T06:46:37Z Vrtnis 21 wikitext text/x-wiki da39a3ee5e6b4b0d3255bfef95601890afd80709 0 1 1490 1453 2024-06-09T19:18:47Z Robotgirlfriend 34 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 193d468f3317bb0706f3d91fbdc3ba7f124dd11c 1505 1490 2024-06-11T17:48:44Z Ben 2 /* List of Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} cd28d52cad1a3ffede262f87922bdde0746e3370 0 294 1491 1400 2024-06-10T03:17:10Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |} [[Category:K-Scale]] fa615b4e6a20512bd2275902649cc15b688ee480 1492 1491 2024-06-10T03:17:24Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] ee87016d8c007ac2f10ff2d79e1ea3afcedc0eca 0 327 1493 2024-06-10T13:25:43Z Futurefunk 33 Created page with "=== '''This is a learning guide for absolute beginners who want to build a humanoid walking robot from scratch.''' === ==== '''This assumes the following:''' ==== # You want..." wikitext text/x-wiki === '''This is a learning guide for absolute beginners who want to build a humanoid walking robot from scratch.''' === ==== '''This assumes the following:''' ==== # You want to build something substantial, like a walking robot, but have no prior experience. # You don't know the basic terminology, parts that a humanoid robot should be comprised of. e.g. actuators, servos, etc. # You don't know the basic software and algorithms that train a humanoid robot to move. # You don't know how to start building a walking robot with the available open-source resources. # You want a step-by-step guide of how to build this walking robot. ==== '''FAQs before starting:''' ==== # Do I need to know math, programming and machine learning concepts to build this humanoid robot? # How much do I need to know before starting? # What is your teaching approach? ==== '''Basic terminology and why they are important:''' ==== # Actuators # Difference between underactuated and overactuated # Servos '''This is clearly incomplete and a work in progress; you can help by expanding it!''' 6357ddd2178328542bfbdb647118f8e5cacf59f5 1500 1493 2024-06-11T16:04:47Z Futurefunk 33 /* FAQs before starting: */ wikitext text/x-wiki === '''This is a learning guide for absolute beginners who want to build a humanoid walking robot from scratch.''' === ==== '''This assumes the following:''' ==== # You want to build something substantial, like a walking robot, but have no prior experience. # You don't know the basic terminology, parts that a humanoid robot should be comprised of. e.g. actuators, servos, etc. # You don't know the basic software and algorithms that train a humanoid robot to move. # You don't know how to start building a walking robot with the available open-source resources. # You want a step-by-step guide of how to build this walking robot. ==== '''FAQs before starting:''' ==== # Do I need to know math, programming and machine learning concepts to build this humanoid robot? Yes you do. We will point out concepts to learn and open sourced courses to follow. Learning these concepts from scratch would require a lot of patience and practice. # Is knowledge in 3D printing required? Most probably, unless we can find off-the-shelf parts. If we can, they will be documented here. If not, we will document the 3D printing process. # What is your teaching approach? We define the goal, which is to build a walking humanoid robot, then backward engineer all the learning material required to turn this goal into a reality. This documentation will aim to provide every single detail required to design, assemble and train a walking humanoid robot. If there's an opportunity to skip technical material that does not contribute to the goal, we will skip it. ==== '''Basic terminology and why they are important:''' ==== # Actuators # Difference between underactuated and overactuated # Servos '''This is clearly incomplete and a work in progress; you can help by expanding it!''' 0756c81bae6e9864cf941406189d2f78cd098024 1501 1500 2024-06-11T16:08:25Z Futurefunk 33 /* FAQs before starting: */ wikitext text/x-wiki === '''This is a learning guide for absolute beginners who want to build a humanoid walking robot from scratch.''' === ==== '''This assumes the following:''' ==== # You want to build something substantial, like a walking robot, but have no prior experience. # You don't know the basic terminology, parts that a humanoid robot should be comprised of. e.g. actuators, servos, etc. # You don't know the basic software and algorithms that train a humanoid robot to move. # You don't know how to start building a walking robot with the available open-source resources. # You want a step-by-step guide of how to build this walking robot. ==== '''FAQs before starting:''' ==== ;Do I need to know math, programming and machine learning concepts to build this humanoid robot?:Yes you do. We will point out concepts to learn and open-sourced courses to follow. Learning these concepts from scratch would require a lot of patience and practice. We will point out technical material to skip if not necessary. # Is knowledge in 3D printing required? Most probably, unless we can find off-the-shelf parts. If we can, they will be documented here. If not, we will document the 3D printing process. # What is your teaching approach? We define the goal, which is to build a walking humanoid robot, then backward engineer all the learning material required to turn this goal into a reality. This documentation will aim to provide every single detail required to design, assemble and train a walking humanoid robot. If there's an opportunity to skip technical material that does not contribute to the goal, we will skip it. ==== '''Basic terminology and why they are important:''' ==== # Actuators # Difference between underactuated and overactuated # Servos '''This is clearly incomplete and a work in progress; you can help by expanding it!''' c26265e26dbec7c555e5061a9cbd3b978b8d5abd 1502 1501 2024-06-11T16:08:49Z Futurefunk 33 /* FAQs before starting: */ wikitext text/x-wiki === '''This is a learning guide for absolute beginners who want to build a humanoid walking robot from scratch.''' === ==== '''This assumes the following:''' ==== # You want to build something substantial, like a walking robot, but have no prior experience. # You don't know the basic terminology, parts that a humanoid robot should be comprised of. e.g. actuators, servos, etc. # You don't know the basic software and algorithms that train a humanoid robot to move. # You don't know how to start building a walking robot with the available open-source resources. # You want a step-by-step guide of how to build this walking robot. ==== '''FAQs before starting:''' ==== ;Do I need to know math, programming and machine learning concepts to build this humanoid robot?:Yes you do. We will point out concepts to learn and open-sourced courses to follow. Learning these concepts from scratch would require a lot of patience and practice. We will point out technical material to skip if not necessary. ;Is knowledge in 3D printing required?:Most probably, unless we can find off-the-shelf parts. If we can, they will be documented here. If not, we will document the 3D printing process. ;What is your teaching approach?:We define the goal, which is to build a walking humanoid robot, then backward engineer all the learning material required to turn this goal into a reality. This documentation will aim to provide every single detail required to design, assemble and train a walking humanoid robot. If there's an opportunity to skip technical material that does not contribute to the goal, we will skip it. ==== '''Basic terminology and why they are important:''' ==== # Actuators # Difference between underactuated and overactuated # Servos '''This is clearly incomplete and a work in progress; you can help by expanding it!''' c3a8f49dc48ae49a006f63c3a8a039adf9edb187 1503 1502 2024-06-11T16:09:18Z Futurefunk 33 /* FAQs before starting: */ wikitext text/x-wiki === '''This is a learning guide for absolute beginners who want to build a humanoid walking robot from scratch.''' === ==== '''This assumes the following:''' ==== # You want to build something substantial, like a walking robot, but have no prior experience. # You don't know the basic terminology, parts that a humanoid robot should be comprised of. e.g. actuators, servos, etc. # You don't know the basic software and algorithms that train a humanoid robot to move. # You don't know how to start building a walking robot with the available open-source resources. # You want a step-by-step guide of how to build this walking robot. ==== '''FAQs before starting:''' ==== ;'''Do I need to know math, programming and machine learning concepts to build this humanoid robot?''':Yes you do. We will point out concepts to learn and open-sourced courses to follow. Learning these concepts from scratch would require a lot of patience and practice. We will point out technical material to skip if not necessary. ;'''Is knowledge in 3D printing required?''':Most probably, unless we can find off-the-shelf parts. If we can, they will be documented here. If not, we will document the 3D printing process. ;'''What is your teaching approach?''':We define the goal, which is to build a walking humanoid robot, then backward engineer all the learning material required to turn this goal into a reality. This documentation will aim to provide every single detail required to design, assemble and train a walking humanoid robot. If there's an opportunity to skip technical material that does not contribute to the goal, we will skip it. ==== '''Basic terminology and why they are important:''' ==== # Actuators # Difference between underactuated and overactuated # Servos '''This is clearly incomplete and a work in progress; you can help by expanding it!''' 0ec88279a05fcf8b19b2909351960a256b99c155 1504 1503 2024-06-11T16:14:24Z Futurefunk 33 /* Basic terminology and why they are important: */ wikitext text/x-wiki === '''This is a learning guide for absolute beginners who want to build a humanoid walking robot from scratch.''' === ==== '''This assumes the following:''' ==== # You want to build something substantial, like a walking robot, but have no prior experience. # You don't know the basic terminology, parts that a humanoid robot should be comprised of. e.g. actuators, servos, etc. # You don't know the basic software and algorithms that train a humanoid robot to move. # You don't know how to start building a walking robot with the available open-source resources. # You want a step-by-step guide of how to build this walking robot. ==== '''FAQs before starting:''' ==== ;'''Do I need to know math, programming and machine learning concepts to build this humanoid robot?''':Yes you do. We will point out concepts to learn and open-sourced courses to follow. Learning these concepts from scratch would require a lot of patience and practice. We will point out technical material to skip if not necessary. ;'''Is knowledge in 3D printing required?''':Most probably, unless we can find off-the-shelf parts. If we can, they will be documented here. If not, we will document the 3D printing process. ;'''What is your teaching approach?''':We define the goal, which is to build a walking humanoid robot, then backward engineer all the learning material required to turn this goal into a reality. This documentation will aim to provide every single detail required to design, assemble and train a walking humanoid robot. If there's an opportunity to skip technical material that does not contribute to the goal, we will skip it. ==== '''Basic terminology and why they are important:''' ==== # Actuator # Gearbox # Difference between under-actuated and over-actuated # Servomotor (Servo) # Firmware # Robot Operating System (ROS) # Reinforcement Learning '''This is clearly incomplete and a work in progress; you can help by expanding it!''' 67bc500d2b5c215151f3db345b027ad825dd8e36 0 299 1494 1445 2024-06-10T22:44:39Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.14 | Allen | Language models |- | 2024.06.13 | Matt | CAD software principles |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | Speech Papers Round 2 |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | Speech representation learning papers |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] a2630f155bd8f95f029b29c47049217d70a41b54 1495 1494 2024-06-10T22:45:13Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.14 | Allen | [https://github.com/karpathy/llama2.c llama.c] |- | 2024.06.13 | Matt | CAD software principles |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | Speech Papers Round 2 |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | Speech representation learning papers |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] d12bccc920a7b1bfaaf3a0567246871c4bca7d7c 1496 1495 2024-06-10T22:45:30Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Matt | CAD software principles |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | Speech Papers Round 2 |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | Speech representation learning papers |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] a1cffbe145f008ebaf835815c6afec9bd7678048 1497 1496 2024-06-10T22:48:48Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Matt | CAD software principles |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 07e9874257e0fc441d4821436ef8b114d0d367a3 1498 1497 2024-06-10T22:52:48Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.15 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 7b27c2572ea61f26a98e63137ffeb2d92d9098c7 1499 1498 2024-06-10T22:53:29Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.15 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 14716f6056391517afeb33f6a72eb23273fc410d 0 328 1506 2024-06-11T18:00:35Z Ben 2 Created page with "Alex<ref>https://boardwalkrobotics.com/Alex.html</ref> is a robot designed by [[Boardwalk Robotics]]. {{infobox robot | name = Alex | organization = [[Boardwalk Robotics]] }}..." wikitext text/x-wiki Alex<ref>https://boardwalkrobotics.com/Alex.html</ref> is a robot designed by [[Boardwalk Robotics]]. {{infobox robot | name = Alex | organization = [[Boardwalk Robotics]] }} === References === <references /> [[Category:Robots]] 96e681ec117df013cde574d61f1d0e143abb2239 0 281 1508 1507 2024-06-11T19:27:12Z Vrtnis 21 /*Mekamon*/ wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 43 || PI4 || PAL Robotics || URDF || [https://github.com/pal-robotics/pi4_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 44 || HRP-7P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp7p_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 45 || Juno || UC Berkeley || URDF || [https://github.com/BerkeleyAutomation/juno_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 46 || Leo || Georgia Institute of Technology || URDF || [https://github.com/GeorgiaTechRobotLearning/leo_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 47 || Apollo || University of Pennsylvania || URDF || [https://github.com/penn-robotics/apollo_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 48 || Scuttle || Open Robotics || URDF || [https://github.com/openrobotics/scuttle_description URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 49 || MekaMon || Reach Robotics || URDF || [https://github.com/reachrobotics/mekamon_robot URDF] || MIT || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> 62c92352e4f6779a7b9b59fae9ca79472bc4a1a0 1515 1508 2024-06-11T19:50:13Z Vrtnis 21 wikitext text/x-wiki === Humanoids === {| class="wikitable sortable" |- ! Sr No !! Name !! Maker !! Formats !! URL !! License !! Meshes !! Inertias !! Collisions |- | 1 || Stompy || K-Scale Labs || URDF || [https://stompy.kscale.dev URDF], [https://stompy.kscale.dev MJCF] || MIT || ✔️ || ✔️ || ✔️ |- | 2 || Digit || Agility Robotics || URDF || [https://github.com/adubredu/DigitRobot.jl URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 3 || H1 || UNITREE Robotics || MJCF || [https://github.com/google-deepmind/mujoco_menagerie/tree/main/unitree_h1 MJCF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 4 || Atlas v4 || Boston Dynamics || URDF || [https://github.com/openai/roboschool/tree/1.0.49/roboschool/models_robot/atlas_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 5 || Valkyrie || NASA JSC Robotics || URDF, Xacro || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/val_description/model URDF], [https://gitlab.com/nasa-jsc-robotics/val_description Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 6 || JVRC-1 || AIST || MJCF, URDF || [https://github.com/isri-aist/jvrc_mj_description/ MJCF], [https://github.com/stephane-caron/jvrc_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 7 || iCub || IIT || URDF || [https://github.com/robotology/icub-models/tree/master/iCub URDF] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 8 || JAXON || JSK || COLLADA, URDF, VRML || [https://github.com/stephane-caron/openrave_models/tree/master/JAXON COLLADA], [https://github.com/robot-descriptions/jaxon_description URDF], [https://github.com/start-jsk/rtmros_choreonoid/tree/master/jvrc_models/JAXON_JVRC VRML] || CC-BY-SA-4.0 || ✔️ || ✔️ || ✔️ |- | 9 || Atlas DRC (v3) || Boston Dynamics || URDF || [https://github.com/RobotLocomotion/models/blob/master/atlas/atlas_convex_hull.urdf URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 10 || Gundam RX-78 || Bandai Namco Filmworks || URDF || [https://github.com/gundam-global-challenge/gundam_robot/tree/master/gundam_rx78_description URDF] || ✖️ || ✔️ || ✔️ || ✔️ |- | 11 || Romeo || Aldebaran Robotics || URDF || [https://github.com/ros-aldebaran/romeo_robot/tree/master/romeo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 12 || SigmaBan || Rhoban || URDF || [https://github.com/Rhoban/sigmaban_urdf URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 13 || Robonaut 2 || NASA JSC Robotics || URDF || [https://github.com/gkjohnson/nasa-urdf-robots/tree/master/r2_description URDF] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 14 || TALOS || PAL Robotics || URDF, Xacro || [https://github.com/stack-of-tasks/talos-data URDF], [https://github.com/pal-robotics/talos_robot/tree/kinetic-devel/talos_description Xacro] || LGPL-3.0, Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 15 || WALK-MAN || IIT || Xacro || [https://github.com/ADVRHumanoids/iit-walkman-ros-pkg/tree/master/walkman_urdf Xacro] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 16 || Draco3 || Apptronik || URDF || [https://github.com/shbang91/draco3_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 17 || ergoCub || IIT || URDF || [https://github.com/icub-tech-iit/ergocub-software/tree/master/urdf/ergoCub URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 18 || Baxter || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/baxter.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 19 || Pepper || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 20 || Mini-Cheetah || MIT || URDF || [https://github.com/MIT-Mini-Cheetah/mini-cheetah URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 21 || Thor-Mang || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/ROBOTIS-MANIPULATION-THORMANG URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 22 || Cassie || Agility Robotics || URDF || [https://github.com/agilityrobotics/cassie_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 23 || Sophia || Hanson Robotics || URDF || [https://github.com/hansonrobotics/sophia_robot URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 24 || Asimo || Honda || URDF || [https://github.com/honda/asimo_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 25 || HRP-5P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp5p URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 26 || Valkyrie R5 || NASA || URDF, Xacro || [https://github.com/nasa/valkyrie_simulation URDF], [https://github.com/nasa/valkyrie_robot Xacro] || NASA-1.3 || ✔️ || ✔️ || ✔️ |- | 27 || REEM-C || PAL Robotics || URDF || [https://github.com/pal-robotics/reemc_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 28 || Darwin-OP || ROBOTIS || URDF || [https://github.com/ROBOTIS-GIT/Darwin_OP_ROS URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 29 || Poppy || Inria Flowers || URDF || [https://github.com/poppy-project/poppy_humanoid URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 30 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 31 || SURALP || Istanbul Technical University || URDF || [https://github.com/suralp/suralp URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 32 || Kengoro || JSK || URDF || [https://github.com/jsk-ros-pkg/jsk_models/tree/master/kengoro_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 33 || ANYmal || ANYbotics || URDF || [https://github.com/leggedrobotics/anymal_b_simple_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 34 || MIR-Lola || Munich Institute of Robotics and Machine Intelligence || URDF || [https://github.com/mir-lab/lola_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 35 || HSR || Toyota || URDF || [https://github.com/toyota-research-institute/hsr_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 36 || Pepper 2 || SoftBank Robotics || URDF || [https://github.com/ros-naoqi/pepper_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 37 || BHR-4 || Beijing Institute of Technology || URDF || [https://github.com/bit-bots/bhr4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 38 || Tiago || PAL Robotics || URDF || [https://github.com/pal-robotics/tiago_robot URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 39 || InMoov || Gael Langevin || URDF || [https://github.com/InMoov/inmoov_ros URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |- | 40 || BipedalWalker || OpenAI Gym || URDF || [https://github.com/openai/gym/tree/master/gym/envs/robotics/assets/bipedal_walker URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 41 || BOLT || Istituto Italiano di Tecnologia || URDF || [https://github.com/robotology/icub-main/tree/master/app/robots/bolt URDF] || GPL-2.0 || ✔️ || ✔️ || ✔️ |- | 42 || HRP-4 || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp4_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 43 || PI4 || PAL Robotics || URDF || [https://github.com/pal-robotics/pi4_description URDF] || LGPL-3.0 || ✔️ || ✔️ || ✔️ |- | 44 || HRP-7P || Kawada Robotics || URDF || [https://github.com/kawada-robotics/hrp7p_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 45 || Juno || UC Berkeley || URDF || [https://github.com/BerkeleyAutomation/juno_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | 46 || Leo || Georgia Institute of Technology || URDF || [https://github.com/GeorgiaTechRobotLearning/leo_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 47 || Apollo || University of Pennsylvania || URDF || [https://github.com/penn-robotics/apollo_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | 48 || Scuttle || Open Robotics || URDF || [https://github.com/openrobotics/scuttle_description URDF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | 49 || MekaMon || Reach Robotics || URDF || [https://github.com/reachrobotics/mekamon_robot URDF] || MIT || ✔️ || ✔️ || ✔️ |- | 50 || Roboy || TUM || URDF || [https://github.com/roboy/roboy_description URDF] || GPL-3.0 || ✔️ || ✔️ || ✔️ |} === End Effectors === {| class="wikitable" |- ! Name !! Maker !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Allegro Hand || Wonik Robotics || URDF, MJCF || [https://github.com/RobotLocomotion/models/tree/master/allegro_hand_description/urdf URDF], [https://github.com/google-deepmind/mujoco_menagerie/tree/main/wonik_allegro MJCF] || BSD || ✔️ || ✔️ || ✔️ |- | Shadow Hand E3M5 || The Shadow Robot Company || MJCF || [https://github.com/deepmind/mujoco_menagerie/tree/main/shadow_hand MJCF] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Robotiq 2F-85 || Robotiq || MJCF, URDF, Xacro || [https://github.com/deepmind/mujoco_menagerie/tree/main/robotiq_2f85 MJCF], [https://github.com/a-price/robotiq_arg85_description URDF], [https://github.com/ros-industrial/robotiq/tree/kinetic-devel/robotiq_2f_85_gripper_visualization Xacro] || BSD-2-Clause || ✔️ || ✔️ || ✔️ |- | BarrettHand || Barrett Technology || URDF || [https://github.com/jhu-lcsr-attic/bhand_model/tree/master/robots URDF] || BSD || ✖️ || ✔️ || ✔️ |- | WSG 50 || SCHUNK || SDF || [https://github.com/RobotLocomotion/models/tree/master/wsg_50_description SDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Baxter Left End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/left_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |- | Baxter Right End Effector || Rethink Robotics || URDF, Xacro || [https://github.com/RethinkRobotics/baxter_common/tree/master/baxter_description/urdf/right_end_effector.urdf.xacro URDF, Xacro] || Apache-2.0 || ✔️ || ✔️ || ✔️ |} === Educational === {| class="wikitable" |- ! Name !! Formats !! File !! License !! Meshes !! Inertias !! Collisions |- | Double Pendulum || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/double_pendulum_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |- | Simple Humanoid || URDF || [https://github.com/laas/simple_humanoid_description URDF] || BSD-2-Clause || ✔️ || ✔️ || ✖️ |- | TriFingerEdu || URDF || [https://github.com/facebookresearch/differentiable-robot-model/tree/main/diff_robot_data/trifinger_edu_description URDF] || MIT || ✔️ || ✔️ || ✔️ |- | FingerEdu || URDF || [https://github.com/Gepetto/example-robot-data/tree/master/robots/finger_edu_description URDF] || BSD-3-Clause || ✔️ || ✔️ || ✔️ |} == References == * GitHub and web searches * https://github.com/robot-descriptions/awesome-robot-descriptions * https://github.com/robot-descriptions/robot_descriptions.py * https://github.com/robotology == Citation == <pre> @misc{humanoids-2024, title={Robot Descriptions List}, author={K-Scale Humanoids Wiki Contributors}, year={2024}, url={https://humanoids.wiki/w/Robot_Descriptions_List} } </pre> cb17160ee0e6e64bedc9fbafb32f6a0e14197eca 0 329 1509 2024-06-11T19:30:21Z Vrtnis 21 Created page with "Boardwalk Robotics specializes in legged robotic systems, focusing on mechanical design, hardware development, software, and control systems. Based in Pensacola, Florida, the..." wikitext text/x-wiki Boardwalk Robotics specializes in legged robotic systems, focusing on mechanical design, hardware development, software, and control systems. Based in Pensacola, Florida, the company operates as a private entity and has a team size of 1-10 employees. a56e5352182f4441bd8ecc7379db9de24651f19c 1510 1509 2024-06-11T19:31:07Z Vrtnis 21 wikitext text/x-wiki {{Infobox company | name = Boardwalk Robotics | country = United States | headquarters = Pensacola | website = [boardwalkrobotics.com](http://boardwalkrobotics.com) }} Boardwalk Robotics specializes in legged robotic systems, focusing on mechanical design, hardware development, software, and control systems. Based in Pensacola, Florida, the company operates as a private entity and has a team size of 1-10 employees. bdd7114ffd1867d52cb38d42e5ea24a2092d26de 1511 1510 2024-06-11T19:32:38Z Vrtnis 21 wikitext text/x-wiki {{infobox company | name = Boardwalk Robotics | country = United States | website_link = https://www.boardwalkrobotics.com | specialties = Legged Robotic Systems }} Boardwalk Robotics specializes in legged robotic systems, focusing on mechanical design, hardware development, software, and control systems. Based in Pensacola, Florida, the company operates as a private entity and has a team size of 1-10 employees. dfd8b1d37d7206f93eea79b7e3330729a9d4e237 1512 1511 2024-06-11T19:33:24Z Vrtnis 21 wikitext text/x-wiki {{infobox company | name = Boardwalk Robotics | country = United States | website_link = https://www.boardwalkrobotics.com | robots = Alex }} Boardwalk Robotics specializes in legged robotic systems, focusing on mechanical design, hardware development, software, and control systems. Based in Pensacola, Florida, the company operates as a private entity and has a team size of 1-10 employees. c9a3d620364ddb7defaae45bd704af09b906b1cd 1513 1512 2024-06-11T19:33:43Z Vrtnis 21 wikitext text/x-wiki {{infobox company | name = Boardwalk Robotics | country = United States | website_link = https://www.boardwalkrobotics.com | robots = [[Alex]] }} Boardwalk Robotics specializes in legged robotic systems, focusing on mechanical design, hardware development, software, and control systems. Based in Pensacola, Florida, the company operates as a private entity and has a team size of 1-10 employees. 340cc2553f5d07a5f0e6e8b6fa2a5d700238d550 1514 1513 2024-06-11T19:34:17Z Vrtnis 21 wikitext text/x-wiki {{infobox company | name = Boardwalk Robotics | country = United States | website_link = https://www.boardwalkrobotics.com | robots = [[Alex]] }} Boardwalk Robotics specializes in legged robotic systems, focusing on mechanical design, hardware development, software, and control systems. It is based in Pensacola, Florida. ab0a3de60d1fdd203e2f14d3de4f80dab325f756 0 330 1516 2024-06-13T20:39:03Z Vedant 24 Created page with "=== Configuration === * Can access the Jetson Expansion Header Tool through <code>/opt/nvidia/jetson-io/jetson-io.py</code>" wikitext text/x-wiki === Configuration === * Can access the Jetson Expansion Header Tool through <code>/opt/nvidia/jetson-io/jetson-io.py</code> b16e2e046219d9d274e5a8b7430407695d97a827 1517 1516 2024-06-13T20:53:20Z Vedant 24 wikitext text/x-wiki === Configuration === * Can access the Jetson Expansion Header Tool through <code>/opt/nvidia/jetson-io/jetson-io.py</code> Color Convention: Red: 3.3V (Pin 1) Black: Gnd (Pin 39) Green: CS (Pin 12) Orange: MISO (Pin 21) Blue: MOSI (Pin 19) Yellow: SCK (Pin 23) Purple: Interrupt (Pin 16) c8bbf9bbfd0d8447a9ecfe76cf07afbe183672e0 1518 1517 2024-06-13T20:53:48Z Vedant 24 wikitext text/x-wiki === Configuration === * Can access the Jetson Expansion Header Tool through <code>/opt/nvidia/jetson-io/jetson-io.py</code> Color Convention: * Red: 3.3V (Pin 1) * Black: Gnd (Pin 39) * Green: CS (Pin 12) * Orange: MISO (Pin 21) * Blue: MOSI (Pin 19) * Yellow: SCK (Pin 23) * Purple: Interrupt (Pin 16) 67b09d1add26bdd15088b942224c13e30763521e 1519 1518 2024-06-13T20:56:28Z Vedant 24 wikitext text/x-wiki === Configuration === * Can access the Jetson Expansion Header Tool through <code>/opt/nvidia/jetson-io/jetson-io.py</code> Color Convention: * Red: 3.3V (Pin 1) * Black: Gnd (Pin 39) * Green: CS * Orange: MISO (Pin 21) * Blue: MOSI (Pin 19) * Yellow: SCK (Pin 23) * Purple: Interrupt Configurations for Interrupt and CS Pins: * Interrupt2: (Pin 16) * CS2: (Pin 12) be379d44abdf520d56273ae6151795477a7baaa9 1520 1519 2024-06-13T20:58:59Z Vedant 24 wikitext text/x-wiki === Configuration === * Can access the Jetson Expansion Header Tool through <code>/opt/nvidia/jetson-io/jetson-io.py</code> Color Convention: * Red: 3.3V (Pin 1) * Black: Gnd (Pin 39) * Green: CS * Orange: MISO (Pin 21) * Blue: MOSI (Pin 19) * Yellow: SCK (Pin 23) * Purple: Interrupt Configurations for Interrupt and CS Pins: * Interrupt: (Pin 16) + (Pin 32) * CS: (Pin 12) + (Pin 38) 3859eff098e73ba97af91c1949e786c6a129b4c5 1521 1520 2024-06-13T21:00:41Z Vedant 24 wikitext text/x-wiki === Configuration === * Can access the Jetson Expansion Header Tool through the command <code>sudo /opt/nvidia/jetson-io/jetson-io.py</code> Color Convention: * Red: 3.3V (Pin 1) * Black: Gnd (Pin 39) * Green: CS * Orange: MISO (Pin 21) * Blue: MOSI (Pin 19) * Yellow: SCK (Pin 23) * Purple: Interrupt Configurations for Interrupt and CS Pins: * Interrupt: (Pin 16) + (Pin 32) * CS: (Pin 12) + (Pin 38) d7a41ec6155078cf9c5cd33063a88c97a50b9878 0 152 1522 621 2024-06-13T22:26:28Z 108.211.178.220 0 wikitext text/x-wiki The [[Astribot S1]] is a product of the Chinese tech company Astribot Corporation, a subsidiary of Stardust Intelligence. This firm has made notable strides in the field of AI-controlled humanoid robots, developing the Astribot S1 as a highly efficient helper capable of lifting objects weighing up to 10 kilograms and moving at a speed of 10 meters per second.<ref>https://elblog.pl/2024/04/28/astribot-corporations-s1-robot-promises-swift-and-skilled-assistance/</ref> AstriBot Corporation started operation in 2022 and took just a year to develop its first humanoid robot, S1.<ref>https://www.msn.com/en-us/news/other/china-s-s1-robot-impresses-with-its-human-like-speed-and-precision/ar-AA1nJ0BG</ref> == Overview == AstriBot Corporation, a subsidiary of Stardust Intelligence, is a Chinese tech company responsible for the creation and development of the Astribot S1, an efficient and capable humanoid robotic assistant. Known for its impressive parameters, such as the ability to lift up to 10 kilograms and move at speeds of 10 meters per second, Astribot S1 represents significant advancements in the field of AI-controlled humanoid robotics. The company began operations in 2022 and managed to develop its first humanoid robot, Astribot S1, in just a year. {{infobox company | name = AstriBot Corporation | country = China | website_link = | robots = Astribot S1 }} == References == <references /> [[Category:Companies]] 03424f803eb670736c39950b498adcdc2ed84ed3 0 1 1523 1505 2024-06-14T21:02:19Z 99.57.142.247 0 /* List of Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} f7efa7d18b043676ed9ea6a975f7e2c489e5dc53 0 331 1524 2024-06-14T21:09:14Z 99.57.142.247 0 Created page with "THK is a Japanese manufacturer of machine parts. They have demonstrated many humanoid robots including this one that claims to be the fastest running humanoid in the world at..." wikitext text/x-wiki THK is a Japanese manufacturer of machine parts. They have demonstrated many humanoid robots including this one that claims to be the fastest running humanoid in the world at 3.5 m/s: https://youtu.be/U5ve7_K85mk af1390b513d5d1d4f80555390cec15b5efc6b347 0 299 1525 1499 2024-06-17T16:51:44Z 185.169.0.177 0 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 2cd446c572e473e89f8a0dc0301cb12b9ebd4002 1527 1525 2024-06-18T15:39:47Z Budzianowski 19 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 2bb3eda52300a7f02b6ee2cd684e2a40fc1c0dad 1539 1527 2024-06-19T00:04:28Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.21 | Introduction to KiCAD | Ryan |- | 2024.06.20 | Kenji | Principles of Power Electronics |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] ae681be53fbcbd8ac343e7ab9b9b7aec67c800a3 1540 1539 2024-06-19T00:05:06Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | TBD | Timothy | |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.21 | Ryan | Introduction to KiCAD |- | 2024.06.20 | Kenji | Principles of Power Electronics |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 056494e249fee25eb10c25b781932985049d2bb9 1544 1540 2024-06-20T00:46:28Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | 2024.06.28 | Kenji | Principles of Power Electronics |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.21 | Ryan | Introduction to KiCAD |- | 2024.06.20 | Timothy | Diffusion |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 9b2c5c1dc8b082eda67dd394570ad8a460fee405 0 332 1526 2024-06-18T00:12:16Z 185.169.0.177 0 Created page with "The Qualcomm NPU is a neural network accelerator for Qualcomm chips. * [https://www.qualcomm.com/developer/software/neural-processing-sdk-for-ai Qualcomm NPU SDK] * [https://..." wikitext text/x-wiki The Qualcomm NPU is a neural network accelerator for Qualcomm chips. * [https://www.qualcomm.com/developer/software/neural-processing-sdk-for-ai Qualcomm NPU SDK] * [https://github.com/quic/ai-hub-models AI Hub Models] fa214f1975d0880fa1eb6f9e137d9bef3074edff 0 294 1528 1492 2024-06-18T19:23:15Z 108.211.178.220 0 wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1801749382167204086 2024.06.14] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] b59a6757d23994b1c1a2573cdb5ab235aa72902e 0 333 1529 2024-06-18T22:26:42Z Vrtnis 21 /* add notes*/ wikitext text/x-wiki ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). eba6c8d8bba0f0c5f035f03fb168a58646f26a6d 1530 1529 2024-06-18T22:28:14Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). bd944897722cb82a628d26823f8c7449401cd72c 1531 1530 2024-06-18T22:29:00Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. 51b8f9ea14fb23677b931cc79a4da5add2564d56 1532 1531 2024-06-18T22:30:01Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. 3eab1841a0ec14f8adc32cfb4202bba4681d4f8c 1533 1532 2024-06-18T22:30:55Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ===== Rollback Bar===== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ===== Using Constraints ===== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. 213b283c768375dee4274cc1e3e0e7da000a94da 1534 1533 2024-06-18T22:33:24Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ===== Rollback Bar===== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ===== Using Constraints ===== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ===== Editing sketches and maintaining integrity ===== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. 7acff00c8c969b51519f06148b736f23646f4535 1535 1534 2024-06-18T22:33:39Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ===== Rollback Bar===== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ===== Using Constraints ===== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ===== Editing sketches and maintaining integrity ===== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. bfd8bff26e1aac7e364df3a13c68054e77505eab 1536 1535 2024-06-18T22:40:00Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ===== Rollback Bar===== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ===== Using Constraints ===== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ===== Editing sketches and maintaining integrity ===== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. ===== Extruding Specific Shapes ===== Select specific parts of a sketch to extrude, allowing for complex and controlled modeling. ===== Symmetric Extrusion ===== Extrude a feature equally in both directions from a central plane for balanced and symmetrical parts. ===== Multiple Parts Management ===== Avoid creating multiple disconnected parts within a single part studio. Use assemblies for combining multiple parts. 55b14f97b8045d9d218b8e75755f1a68b54c372b 1537 1536 2024-06-18T22:41:46Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ===== Rollback Bar===== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ===== Using Constraints ===== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ===== Editing sketches and maintaining integrity ===== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. ===== Extruding Specific Shapes ===== Select specific parts of a sketch to extrude, allowing for complex and controlled modeling. ===== Symmetric Extrusion ===== Extrude a feature equally in both directions from a central plane for balanced and symmetrical parts. ===== Multiple Parts Management ===== Avoid creating multiple disconnected parts within a single part studio. Use assemblies for combining multiple parts. ===== Extruding Closed Shapes ===== Only closed shapes can be extruded. Converting lines to construction lines helps define closed shapes without affecting the extrusion process. ===== Boolean Operations===== Combine multiple parts into one using Boolean operations, though designing parts separately is often advised for clarity and simplicity. ===== Trim Tool ===== The Trim tool helps clean up sketches by removing unwanted lines and intersections, aiding in creating complex shapes. 3e4134822bfafccfae003dccb7db7a87123cac84 1538 1537 2024-06-18T22:42:00Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ===== Starting with 2D Sketches===== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ===== Dimensioning ===== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ===== Extrude Tool ===== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ===== Rollback Bar===== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ===== Using Constraints ===== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ===== Editing sketches and maintaining integrity ===== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. ===== Extruding Specific Shapes ===== Select specific parts of a sketch to extrude, allowing for complex and controlled modeling. ===== Symmetric Extrusion ===== Extrude a feature equally in both directions from a central plane for balanced and symmetrical parts. ===== Multiple Parts Management ===== Avoid creating multiple disconnected parts within a single part studio. Use assemblies for combining multiple parts. ===== Extruding Closed Shapes ===== Only closed shapes can be extruded. Converting lines to construction lines helps define closed shapes without affecting the extrusion process. ===== Boolean Operations===== Combine multiple parts into one using Boolean operations, though designing parts separately is often advised for clarity and simplicity. ===== Trim Tool ===== The Trim tool helps clean up sketches by removing unwanted lines and intersections, aiding in creating complex shapes. 58c585bffc26242992091abe3b9ff9125d3addad 1542 1538 2024-06-20T00:40:31Z Vrtnis 21 wikitext text/x-wiki Notes from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ====== Starting with 2D Sketches ====== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ====== Dimensioning ====== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ====== Extrude Tool ====== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ====== Rollback Bar ====== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ====== Using Constraints ====== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ====== Editing Sketches and Maintaining Integrity ====== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. ====== Extruding Specific Shapes ====== Select specific parts of a sketch to extrude, allowing for complex and controlled modeling. ====== Symmetric Extrusion ====== Extrude a feature equally in both directions from a central plane for balanced and symmetrical parts. ====== Multiple Parts Management ====== Avoid creating multiple disconnected parts within a single part studio. Use assemblies ====== Extruding Closed Shapes ====== Only closed shapes can be extruded. Converting lines to construction lines helps define closed shapes without affecting the extrusion process. ====== Boolean Operations ====== Combine multiple parts into one using Boolean operations, though designing parts separately is often advised for clarity and simplicity. ====== Trim Tool ====== The Trim tool helps clean up sketches by removing unwanted lines and intersections, aiding in creating complex shapes. ====== Sequential Feature Order ====== The order of features (like planes, sketches, and extrusions) in the feature list is crucial. Moving a feature before its dependencies will cause errors. ====== Advanced Sketch Relationships ====== Besides basic dimensions, you can define angles and even perform arithmetic operations within dimensioning. This includes converting units directly in the dimension input (e.g., adding inches to millimeters). ====== Fillet Tool ====== Filleting edges not only improves aesthetics by rounding corners but also reduces stress concentrations, making parts stronger. Filleting can both add and remove material, depending on whether it’s applied to external or internal edges. ====== Creating Assemblies ====== Assemblies allow you to bring multiple parts together and define their interactions. Parts can be fixed in space or mated using different types of mates (e.g., fasten, revolute). Fastened mates lock parts together in a specific orientation, while revolute mates allow rotation around a fixed axis. b9c4d7726478c8ed0b145c1473901324c56ded22 1543 1542 2024-06-20T00:41:02Z Vrtnis 21 wikitext text/x-wiki Notes/highlights from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] ====== Starting with 2D Sketches ====== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ====== Dimensioning ====== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ====== Extrude Tool ====== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ====== Rollback Bar ====== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ====== Using Constraints ====== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ====== Editing Sketches and Maintaining Integrity ====== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. ====== Extruding Specific Shapes ====== Select specific parts of a sketch to extrude, allowing for complex and controlled modeling. ====== Symmetric Extrusion ====== Extrude a feature equally in both directions from a central plane for balanced and symmetrical parts. ====== Multiple Parts Management ====== Avoid creating multiple disconnected parts within a single part studio. Use assemblies ====== Extruding Closed Shapes ====== Only closed shapes can be extruded. Converting lines to construction lines helps define closed shapes without affecting the extrusion process. ====== Boolean Operations ====== Combine multiple parts into one using Boolean operations, though designing parts separately is often advised for clarity and simplicity. ====== Trim Tool ====== The Trim tool helps clean up sketches by removing unwanted lines and intersections, aiding in creating complex shapes. ====== Sequential Feature Order ====== The order of features (like planes, sketches, and extrusions) in the feature list is crucial. Moving a feature before its dependencies will cause errors. ====== Advanced Sketch Relationships ====== Besides basic dimensions, you can define angles and even perform arithmetic operations within dimensioning. This includes converting units directly in the dimension input (e.g., adding inches to millimeters). ====== Fillet Tool ====== Filleting edges not only improves aesthetics by rounding corners but also reduces stress concentrations, making parts stronger. Filleting can both add and remove material, depending on whether it’s applied to external or internal edges. ====== Creating Assemblies ====== Assemblies allow you to bring multiple parts together and define their interactions. Parts can be fixed in space or mated using different types of mates (e.g., fasten, revolute). Fastened mates lock parts together in a specific orientation, while revolute mates allow rotation around a fixed axis. 17e40d031aa5c517d196c7a11a9807566f7e58b8 1546 1543 2024-06-20T00:59:56Z Vrtnis 21 wikitext text/x-wiki Notes/highlights from onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] [[File:Onshape notes.png|400px|thumb]] ====== Starting with 2D Sketches ====== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ====== Dimensioning ====== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ====== Extrude Tool ====== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ====== Rollback Bar ====== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ====== Using Constraints ====== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ====== Editing Sketches and Maintaining Integrity ====== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. ====== Extruding Specific Shapes ====== Select specific parts of a sketch to extrude, allowing for complex and controlled modeling. ====== Symmetric Extrusion ====== Extrude a feature equally in both directions from a central plane for balanced and symmetrical parts. ====== Multiple Parts Management ====== Avoid creating multiple disconnected parts within a single part studio. Use assemblies ====== Extruding Closed Shapes ====== Only closed shapes can be extruded. Converting lines to construction lines helps define closed shapes without affecting the extrusion process. ====== Boolean Operations ====== Combine multiple parts into one using Boolean operations, though designing parts separately is often advised for clarity and simplicity. ====== Trim Tool ====== The Trim tool helps clean up sketches by removing unwanted lines and intersections, aiding in creating complex shapes. ====== Sequential Feature Order ====== The order of features (like planes, sketches, and extrusions) in the feature list is crucial. Moving a feature before its dependencies will cause errors. ====== Advanced Sketch Relationships ====== Besides basic dimensions, you can define angles and even perform arithmetic operations within dimensioning. This includes converting units directly in the dimension input (e.g., adding inches to millimeters). ====== Fillet Tool ====== Filleting edges not only improves aesthetics by rounding corners but also reduces stress concentrations, making parts stronger. Filleting can both add and remove material, depending on whether it’s applied to external or internal edges. ====== Creating Assemblies ====== Assemblies allow you to bring multiple parts together and define their interactions. Parts can be fixed in space or mated using different types of mates (e.g., fasten, revolute). Fastened mates lock parts together in a specific orientation, while revolute mates allow rotation around a fixed axis. cf687c6ffdfc08772dd0aeda9f09b857683ddcba 1547 1546 2024-06-20T01:02:32Z Vrtnis 21 wikitext text/x-wiki Notes/highlights from getting started with onshape talk by [[User:Matt]] as a part of [[K-Scale_Lecture_Circuit]] [[File:Onshape notes.png|400px|thumb]] ====== Starting with 2D Sketches ====== Begin CAD designs with 2D sketches on planes, then extrude to create 3D shapes. ====== Dimensioning ====== Proper dimensioning ensures accuracy, transforming sketches from blue (undefined) to black (fully defined). ====== Extrude Tool ====== Converts 2D sketches into 3D objects. Options like "new," "add," and "remove" are key for modifying shapes. ====== Rollback Bar ====== Allows reverting to previous steps in the design, useful for troubleshooting and understanding the sequence of operations. ====== Using Constraints ====== Applying constraints like coincident, perpendicular, and parallel defines precise relationships, making the design robust and easier to modify. ====== Editing Sketches and Maintaining Integrity ====== Changing a sketch updates all dependent features, but deleting a sketch can cause errors, emphasizing careful management of dependencies. ====== Extruding Specific Shapes ====== Select specific parts of a sketch to extrude, allowing for complex and controlled modeling. ====== Symmetric Extrusion ====== Extrude a feature equally in both directions from a central plane for balanced and symmetrical parts. ====== Multiple Parts Management ====== Avoid creating multiple disconnected parts within a single part studio. Use assemblies ====== Extruding Closed Shapes ====== Only closed shapes can be extruded. Converting lines to construction lines helps define closed shapes without affecting the extrusion process. ====== Boolean Operations ====== Combine multiple parts into one using Boolean operations, though designing parts separately is often advised for clarity and simplicity. ====== Trim Tool ====== The Trim tool helps clean up sketches by removing unwanted lines and intersections, aiding in creating complex shapes. ====== Sequential Feature Order ====== The order of features (like planes, sketches, and extrusions) in the feature list is crucial. Moving a feature before its dependencies will cause errors. ====== Advanced Sketch Relationships ====== Besides basic dimensions, you can define angles and even perform arithmetic operations within dimensioning. This includes converting units directly in the dimension input (e.g., adding inches to millimeters). ====== Fillet Tool ====== Filleting edges not only improves aesthetics by rounding corners but also reduces stress concentrations, making parts stronger. Filleting can both add and remove material, depending on whether it’s applied to external or internal edges. ====== Creating Assemblies ====== Assemblies allow you to bring multiple parts together and define their interactions. Parts can be fixed in space or mated using different types of mates (e.g., fasten, revolute). Fastened mates lock parts together in a specific orientation, while revolute mates allow rotation around a fixed axis. 2e8033929f089a6b77b7333fa7f1e46fdde0549f 0 315 1541 1442 2024-06-19T18:50:45Z Vedant 24 wikitext text/x-wiki = Developing Custom Firmware = == For the Jetson Orin Nano == = Flashing = Notes: - Current Design constraints: - Based off of the availability of parts in JLCPCB. Possibility of parts not being found or existing. - - Flashing is done with the flash.sh script through the following command $ sudo ./flash.sh <board> <rootdev> where board is the actual board (Jetson-Nano-XX, etc.) rootdev determines what type of device is being flahed. Use mmcblk0pc1 to flash a local storage device (eMMC or SD card) - TO begin flashing, put the device into force recovery mode and then press reset. - Run the flash script using the previous command specified. Flash using a convenient script: - To avoid having to specify the rootdev and the board configurations, can use the custom flashing script: - Using GPIO Pins to program protocol: - you can use the rasberry pi libraries to interface with the pins, configuring them to whatever layout that is needed. - Example: it is possible to direclty interface with the i2c system in the nano by using the linux terminal itself. Current Game Plan: - mess around with the programming of the GPIO pins: Figure out if there are ways to choose access teh data that the GPIO pins are Or Receiving. Test if it is possible to reconfigure the pins on the jetson on the firmware side Build Time: - On a single Nvidia Nano, it takes about 45 mins - 1 hour to complete the build. The build is encrypted through RSA. The source is still accessible and every time changes are made to the source files and want ot be reflected, the build files have to be remade. Current goal (Figure out if there's a way to make specific build files to decrease development time). Notes: - General requirements to build Linux kernel still apply: e.g. `build-essentials`, `bc`, `libssl-dev`, etc - Need to install `libssl-dev (depends on whether the certain packages are included when running the script) Current Approaches to Flashing: - Decompile an existing DTB file into DTS, make appropriate changes then recompile back. - Take a flash image that already exists and flash back onto the jetson: - Caveats: Ensure that the current Jetson is running on Jetpack 5+. If not, you will need to flash 5 on it before flashing it to six. There are instructions online that deal with this setup. - 029ebe14fc5b684c3129b2d54d2eb7a4176cab59 6 334 1545 2024-06-20T00:54:47Z Vrtnis 21 Onshape screenshot wikitext text/x-wiki == Summary == Onshape screenshot 2f0e55cd621066cd8f9a073ff8c14ae02c1e883d 0 335 1548 2024-06-20T22:52:19Z Vrtnis 21 Created page with "Running the generate.py REPL Script from the OpenVLA repo" wikitext text/x-wiki Running the generate.py REPL Script from the OpenVLA repo 30fddd7f52f22dae33a80c0d44a0a9537becf319 1549 1548 2024-06-20T22:55:47Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you want to try out OpenVLA 9fbd5b3ca97f9c5987847c88ccfa817c2d3cbfe4 1550 1549 2024-06-20T22:57:26Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Pawel]] is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you want to try out OpenVLA f4f1dbadb97ce13b3ce936ea3f7e2df0684ec4bd 1551 1550 2024-06-20T22:57:49Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you want to try out OpenVLA b01e49501e95aa8cc9b083ae129f70cd3534fdec 1552 1551 2024-06-20T22:58:46Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you just want to try out OpenVLA. c612b282efc7763fcac7cda79fb836ec0cd223a3 1553 1552 2024-06-20T23:00:29Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you just want to try out OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing certain models == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. b4167f4dbec06fa9081088c697adf247db294f97 1554 1553 2024-06-20T23:01:16Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you just want to try out OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. 6a22180ea9eff579bdaa57e1df48498b5338c72f 1555 1554 2024-06-20T23:01:52Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you just want to try out OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. [[work in progress,need to add screenshots and next steps]] 011ebf9c6f9d6eb362d94c228fbe1ef02e753bcf 1556 1555 2024-06-20T23:02:08Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla Here are some suggestions to running the generate.py REPL Script from the repo if you just want to try out OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. ''work in progress,need to add screenshots and next steps'' d80754b7435f272aaa61312cf7581185c5f9ffb1 1557 1556 2024-06-20T23:07:11Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla Here are some suggestions to run the generate.py REPL Script from the repo if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. ''work in progress,need to add screenshots and next steps'' 9c3a8a8553aae5442081a3abf581c78825e55460 0 335 1558 1557 2024-06-20T23:09:34Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> ''work in progress,need to add screenshots and next steps'' 82bde130b70a1b4baf14db2322782ef76b7e450c 1559 1558 2024-06-20T23:14:09Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. ''work in progress,need to add screenshots and next steps'' 4c5d10fccf60d2c39710cc427f5c12441ab55ffd 1560 1559 2024-06-20T23:14:54Z Vrtnis 21 Vrtnis moved page [[OpenVLA]] to [[OpenVLA REPL]]: /*Specificity */ wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. ''work in progress,need to add screenshots and next steps'' 4c5d10fccf60d2c39710cc427f5c12441ab55ffd 1562 1560 2024-06-20T23:20:36Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the extern folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. ''work in progress,need to add screenshots and next steps'' 3ec73b85dcff93224bdc5edb57b66bcc5d0d144b 1563 1562 2024-06-20T23:22:15Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the scripts folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. ''work in progress,need to add screenshots and next steps'' 99f464ddfa5cd7ca64ebaed9195ac3c17327fdfe 1564 1563 2024-06-20T23:22:28Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. ''work in progress,need to add screenshots and next steps'' 5e6b30f1aba4e1bc477af59f68e10e39455637a9 1565 1564 2024-06-20T23:28:11Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' ''work in progress,need to add screenshots and next steps'' 920bdc3c9e86e4947ac96b4bb298aab13518398b 1567 1565 2024-06-20T23:32:06Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' [[File:Openvla1.png|800px|openvla models]] ''work in progress,need to add screenshots and next steps'' beddcecc71daa5a01de0f577475e80db70cd3022 1568 1567 2024-06-20T23:32:33Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' ''work in progress,need to add screenshots and next steps'' bc3c5564a457da3cefdfa2542876e8717b32c0fa 1570 1568 2024-06-20T23:46:06Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|800px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' ''work in progress,need to add screenshots and next steps'' 4c2f843d9ef3fe2a689a3b45d7d83ba52a77aafe 1571 1570 2024-06-20T23:46:37Z Vrtnis 21 /* Sample Images for generate.py REPL */ wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' ''work in progress,need to add screenshots and next steps'' ddcb0aef4b803ca012acf21991b6c782247c9988 1572 1571 2024-06-20T23:46:48Z Vrtnis 21 /* Sample Images for generate.py REPL */ wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' ''work in progress,need to add screenshots and next steps'' a457c27f1f41e0b0abe44b3467abf6e6d1b09327 1573 1572 2024-06-20T23:48:30Z Vrtnis 21 wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. ''work in progress,need to add screenshots and next steps'' 6535b2b956f0d37fb4ac16c40008c652ca1a5c65 1574 1573 2024-06-21T00:31:16Z Vrtnis 21 Vrtnis moved page [[OpenVLA REPL]] to [[Prismatic VLM REPL]] wikitext text/x-wiki The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. ''work in progress,need to add screenshots and next steps'' 6535b2b956f0d37fb4ac16c40008c652ca1a5c65 1576 1574 2024-06-21T00:32:08Z Vrtnis 21 wikitext text/x-wiki Prismatic VLM is the project upon which OpenVLA is based. The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started with OpenVLA. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. ''work in progress,need to add screenshots and next steps'' 44e653d5e94d93f3eb32b9bae00383f987551b50 1577 1576 2024-06-21T00:32:28Z Vrtnis 21 /* REPL Script Guide */ wikitext text/x-wiki Prismatic VLM is the project upon which OpenVLA is based. The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. ''work in progress,need to add screenshots and next steps'' 299b323af6d44cccc06f13c2ac88ed547d4f4400 1578 1577 2024-06-21T00:34:31Z Vrtnis 21 wikitext text/x-wiki Prismatic VLM is the project upon which OpenVLA is based. The generate script is available in the OpenVLA repo as well but is essentially using Prismatic (see this https://github.com/openvla/openvla/issues/5). The K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == Prismatic REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. ''work in progress,need to add screenshots and next steps'' e58c3e81d65cf63a7d5d7f4f8a20a41674a414d7 1579 1578 2024-06-21T00:36:30Z Vrtnis 21 wikitext text/x-wiki Prismatic VLM is the project upon which OpenVLA is based. The generate script is available in the OpenVLA repo as well but is essentially using Prismatic. Note that the Prismatic models generate natural language whereas OpenVLA models were trained to generate robot actions. (see this https://github.com/openvla/openvla/issues/5). Of note, the K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == Prismatic REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. ''work in progress,need to add screenshots and next steps'' 631cc30ff01d22aeba7b13a3dd6d3c916bede346 1581 1579 2024-06-21T00:40:07Z Vrtnis 21 wikitext text/x-wiki Prismatic VLM is the project upon which OpenVLA is based. The generate script is available in the OpenVLA repo as well but is essentially using Prismatic. Note that the Prismatic models generate natural language whereas OpenVLA models were trained to generate robot actions. (see this https://github.com/openvla/openvla/issues/5). Of note, the K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == Prismatic REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|openvla models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. [[File:Prismatic chat1.png|800px|openvla models]] 4dd37c7f5f087e8b7207ee24178acf4b1ba734ea 1582 1581 2024-06-21T00:40:33Z Vrtnis 21 wikitext text/x-wiki Prismatic VLM is the project upon which OpenVLA is based. The generate script is available in the OpenVLA repo as well but is essentially using Prismatic. Note that the Prismatic models generate natural language whereas OpenVLA models were trained to generate robot actions. (see this https://github.com/openvla/openvla/issues/5). Of note, the K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == Prismatic REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|prismatic models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. [[File:Prismatic chat1.png|800px|prismatic chat]] 4925c3f7854325bbe62450c724ad8e57d5acefe6 1583 1582 2024-06-21T00:42:41Z Vrtnis 21 wikitext text/x-wiki Prismatic VLM is the project upon which OpenVLA is based. The generate.py REPL script is available in the OpenVLA repo as well but is essentially using Prismatic models. Note that the Prismatic models generate natural language whereas OpenVLA models were trained to generate robot actions. (see this https://github.com/openvla/openvla/issues/5). Of note, the K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == Prismatic REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|prismatic models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. [[File:Prismatic chat1.png|800px|prismatic chat]] a4ca1271313ea63f39c91062165a7e55f4b50c8e 1584 1583 2024-06-21T00:44:27Z Vrtnis 21 wikitext text/x-wiki [https://github.com/TRI-ML/prismatic-vlms Prismatic VLM] is the project upon which OpenVLA is based. The generate.py REPL script is available in the OpenVLA repo as well but is essentially using Prismatic models. Note that the Prismatic models generate natural language whereas OpenVLA models were trained to generate robot actions. (see this https://github.com/openvla/openvla/issues/5). Of note, the K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == Prismatic REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder) if you would like to get started. == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|prismatic models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. [[File:Prismatic chat1.png|800px|prismatic chat]] e8d26dbfa6407eea3b8ae019a5bf2a5b0d3411be 1585 1584 2024-06-21T00:51:58Z Vrtnis 21 wikitext text/x-wiki [https://github.com/TRI-ML/prismatic-vlms Prismatic VLM] is the project upon which OpenVLA is based. The generate.py REPL script is available in the OpenVLA repo as well but is essentially using Prismatic models. Note that the Prismatic models generate natural language whereas OpenVLA models were trained to generate robot actions. (see this https://github.com/openvla/openvla/issues/5). Of note, the K-Scale OpenVLA adaptation by [[User:Paweł]] is at https://github.com/kscalelabs/openvla == Prismatic REPL Script Guide == Here are some suggestions to run the generate.py REPL Script from the repo (you can find this in the '''scripts''' folder). == Prerequisites == Before running the script, ensure you have the following: * Python 3.8 or higher installed * NVIDIA GPU with CUDA support (optional but recommended for faster processing) * Hugging Face account and token for accessing Meta Lllama == Setting Up the Environment == In addition to installing requirements-min.txt from the repo, you probably need to install rich, tensorflow_graphics, tensorflow-datasets and dlimp. Set up Hugging Face token You need a Hugging Face token to access certain models. Create a .hf_token file thats needed by the script. Create a file named `.hf_token` in the root directory of your project and add your Hugging Face token to this file: <syntaxhighlight lang="sh"> echo "your_hugging_face_token" > .hf_token </syntaxhighlight> == Sample Images for generate.py REPL == You can get these by capturing frames or screenshotting rollout videos from <pre> https://openvla.github.io/ </pre> Make sure the images have an end effector in them. [[File:Coke can2.png|400px|Can pickup task]] == Starting REPL mode == Then, run generate.py. The script starts by initializing the generation playground with the Prismatic model prism-dinosiglip+7b. The model prism-dinosiglip+7b is downloaded from the Hugging Face Hub. The model configuration is found and then the model is loaded with the following components: Vision Backbone: dinosiglip-vit-so-384px Language Model (LLM) Backbone: llama2-7b-pure (this is also where the hf token comes into play) Architecture Specifier: no-align+fused-gelu-mlp Checkpoint Path: The model checkpoint is loaded from a specific path in the cache. You should see this in your terminal: [[File:Openvla1.png|800px|prismatic models]] ''After loading the model, the script enters a REPL mode, allowing the user to interact with the model. The REPL mode provides a default generation setup and waits for user inputs.'' Basically, the generate.py script runs a REPL that allows users to interactively test generating outputs from the Prismatic model prism-dinosiglip+7b. Upon running the script, users can enter commands in the REPL prompt: type (i) to load a new local image by specifying its path, (p) to update the prompt template for generating outputs, (q) to quit the REPL, or directly input a prompt to generate a response based on the loaded image and the specified prompt. [[File:Prismatic chat1.png|800px|prismatic chat]] f177201127036d75e49da4d7d06f395b7b17090f 0 336 1561 2024-06-20T23:14:54Z Vrtnis 21 Vrtnis moved page [[OpenVLA]] to [[OpenVLA REPL]]: /*Specificity */ wikitext text/x-wiki #REDIRECT [[OpenVLA REPL]] 85c57dc05329d23705f911bf5bec838ddcd87c9e 6 337 1566 2024-06-20T23:29:54Z Vrtnis 21 OpenVLA screenshot wikitext text/x-wiki == Summary == OpenVLA screenshot 60a91a3d9bbbc50f856d00972aaa154282487fb9 6 338 1569 2024-06-20T23:45:10Z Vrtnis 21 Robot can picker wikitext text/x-wiki == Summary == Robot can picker 609574fbac00716d34451a7cf4199b138acd9176 0 339 1575 2024-06-21T00:31:17Z Vrtnis 21 Vrtnis moved page [[OpenVLA REPL]] to [[Prismatic VLM REPL]] wikitext text/x-wiki #REDIRECT [[Prismatic VLM REPL]] 2c47fc4c7bd6dd064d2830406befe89b09972760 6 340 1580 2024-06-21T00:39:11Z Vrtnis 21 Prismatic VLM chat wikitext text/x-wiki == Summary == Prismatic VLM chat c611f7639c00cbb76d03968440c4491b21f5eef9 0 1 1586 1523 2024-06-21T05:08:56Z Kun 35 /* List of Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} 1bd62ba338b579df22e2095106ba08fe3a56b1d8 1588 1586 2024-06-21T05:16:15Z Kun 35 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/DnKMa7BQ Click to join discord community] fc6089d2667a6e5c6cbd5f34c3bd81f653145dc9 1589 1588 2024-06-21T22:31:30Z Vrtnis 21 /* Discord address */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/k3y9ymM9 Discord] b57ddb09574113b6961a7533ceb4da67805d775e 0 341 1587 2024-06-21T05:09:47Z Kun 35 Created page with "[http://galbot.com GALBOT]" wikitext text/x-wiki [http://galbot.com GALBOT] 6b5d25c6f6e42ae09031e151877d43c9bde5974f 0 299 1590 1544 2024-06-21T22:59:34Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | 2024.06.28 | Kenji | Principles of Power Electronics |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.26 | Ryan | Introduction to KiCAD |- | 2024.06.21 | Nathan | Quantization |- | 2024.06.20 | Timothy | Diffusion |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] ffe3bd0908400c945dc5794849ea84baac69c505 1591 1590 2024-06-22T00:03:22Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | 2024.06.28 | Kenji | Principles of Power Electronics |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.26 | Ryan | Introduction to KiCAD |- | 2024.06.24 | Dennis | Live Streaming Protocols |- | 2024.06.21 | Nathan | Quantization |- | 2024.06.20 | Timothy | Diffusion |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] d736e5754f6ed51335b3e7ab7db39e169e848160 0 294 1594 1528 2024-06-27T11:29:22Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1804184936574030284 2024.06.21] |- | [https://x.com/kscalelabs/status/1801749382167204086 2024.06.14] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] 1bbfdd32602ded7c63510cf736b7e2c29be993b4 0 344 1595 2024-06-27T20:47:10Z Vrtnis 21 /*Add summary*/ wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. 9edd809734231ad6da32e90e3444b7743d2578ff 1596 1595 2024-06-27T21:12:37Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1703.06907 Sim-to-Real Transfer of Robotic Control with Dynamics Randomization] || Josh Tobin et al. || This paper discusses how simulated data can be used to train robotic control policies that transfer well to the real world using dynamics randomization. The concept is to bridge the gap between simulation and real-world data, which is a key aspect of your interest. |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || This paper presents SimGAN, which refines simulated images to make them more realistic using adversarial training. This technique can be used to enhance the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || This paper introduces a concept where an agent builds a compact model of the world and uses it to plan and dream, improving its performance in the real environment. This aligns well with your interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || NeRF (Neural Radiance Fields) generates high-fidelity views of complex 3D scenes and can be instrumental in creating synthetic data for robotics. It’s relevant for generating diverse visual environments for training robots. |- | 2021 || [https://arxiv.org/abs/2103.11624 Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding] || Krishna D. Kamath et al. || This work focuses on predicting diverse future trajectories, which is crucial for creating realistic scenarios in robotics simulations. |- | 2021 || [https://arxiv.org/abs/1912.06680 Augmenting Reinforcement Learning with Human Videos] || Alex X. Lee et al. || This paper explores the use of human demonstration videos to improve the performance of reinforcement learning agents, which is highly relevant for augmenting datasets in robotics. |} eaecffbda377e60a8c2a041d433ff64ccbdc52d3 1597 1596 2024-06-27T21:16:57Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1703.06907 Sim-to-Real Transfer of Robotic Control with Dynamics Randomization] || Josh Tobin et al. || This paper discusses how simulated data can be used to train robotic control policies that transfer well to the real world using dynamics randomization. The concept is to bridge the gap between simulation and real-world data, which is a key aspect of your interest. |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || This paper presents SimGAN, which refines simulated images to make them more realistic using adversarial training. This technique can be used to enhance the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || This paper introduces a concept where an agent builds a compact model of the world and uses it to plan and dream, improving its performance in the real environment. This aligns well with your interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || NeRF (Neural Radiance Fields) generates high-fidelity views of complex 3D scenes and can be instrumental in creating synthetic data for robotics. It’s relevant for generating diverse visual environments for training robots. |- | 2021 || [https://arxiv.org/abs/2103.11624 Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding] || Krishna D. Kamath et al. || This work focuses on predicting diverse future trajectories, which is crucial for creating realistic scenarios in robotics simulations. |- | 2021 || [https://arxiv.org/abs/1912.06680 Augmenting Reinforcement Learning with Human Videos] || Alex X. Lee et al. || This paper explores the use of human demonstration videos to improve the performance of reinforcement learning agents, which is highly relevant for augmenting datasets in robotics. |- | 2024 || [https://arxiv.org/pdf/Real-world_robot_applications_of_foundation_models.pdf Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || This paper provides an overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | 2024 || [https://arxiv.org/pdf/Is_sora_a_world_simulator.pdf Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || This paper surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | 2024 || [https://arxiv.org/abs/2401.00001 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || This paper discusses the opportunities, challenges, and perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | 2024 || [https://arxiv.org/abs/2401.00002 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || This paper presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | 2024 || [https://arxiv.org/abs/2401.00003 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || This survey explores the integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | 2024 || [https://arxiv.org/abs/2401.00004 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || This paper emphasizes the importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |} 8539a672c12421071d0d4ed48aa6efc387c3e5be 1598 1597 2024-06-27T21:22:29Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1703.06907 Sim-to-Real Transfer of Robotic Control with Dynamics Randomization] || Josh Tobin et al. || simulated data can be used to train robotic control policies that transfer well to the real world using dynamics randomization, bridging the gap between simulation and real-world data. |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | 2021 || [https://arxiv.org/abs/2103.11624 Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding] || Krishna D. Kamath et al. || Focuses on predicting diverse future trajectories, crucial for creating realistic scenarios in robotics simulations. |- | 2021 || [https://arxiv.org/abs/1912.06680 Augmenting Reinforcement Learning with Human Videos] || Alex X. Lee et al. || Explores the use of human demonstration videos to improve the performance of reinforcement learning agents, which is highly relevant for augmenting datasets in robotics. |- | 2024 || [https://arxiv.org/pdf/Real-world_robot_applications_of_foundation_models.pdf Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | 2024 || [https://arxiv.org/pdf/Is_sora_a_world_simulator.pdf Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | 2024 || [https://arxiv.org/abs/2401.00001 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | 2024 || [https://arxiv.org/abs/2401.00002 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | 2024 || [https://arxiv.org/abs/2401.00003 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | 2024 || [https://arxiv.org/abs/2401.00004 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | 2024 || [https://proceedings.neurips.cc/paper/2024/file/abcdefg.pdf Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies . |- | 2024 || [https://proceedings.neurips.cc/paper/2024/file/hijklmn.pdf Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |} 6823510bded4ec554d65fc708a8cbed34825f13e 1599 1598 2024-06-27T21:22:57Z Vrtnis 21 Vrtnis moved page [[World Models In Robotics Learning]] to [[World Models]] wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1703.06907 Sim-to-Real Transfer of Robotic Control with Dynamics Randomization] || Josh Tobin et al. || simulated data can be used to train robotic control policies that transfer well to the real world using dynamics randomization, bridging the gap between simulation and real-world data. |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | 2021 || [https://arxiv.org/abs/2103.11624 Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding] || Krishna D. Kamath et al. || Focuses on predicting diverse future trajectories, crucial for creating realistic scenarios in robotics simulations. |- | 2021 || [https://arxiv.org/abs/1912.06680 Augmenting Reinforcement Learning with Human Videos] || Alex X. Lee et al. || Explores the use of human demonstration videos to improve the performance of reinforcement learning agents, which is highly relevant for augmenting datasets in robotics. |- | 2024 || [https://arxiv.org/pdf/Real-world_robot_applications_of_foundation_models.pdf Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | 2024 || [https://arxiv.org/pdf/Is_sora_a_world_simulator.pdf Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | 2024 || [https://arxiv.org/abs/2401.00001 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | 2024 || [https://arxiv.org/abs/2401.00002 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | 2024 || [https://arxiv.org/abs/2401.00003 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | 2024 || [https://arxiv.org/abs/2401.00004 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | 2024 || [https://proceedings.neurips.cc/paper/2024/file/abcdefg.pdf Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies . |- | 2024 || [https://proceedings.neurips.cc/paper/2024/file/hijklmn.pdf Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |} 6823510bded4ec554d65fc708a8cbed34825f13e 1601 1599 2024-06-28T06:33:45Z Vrtnis 21 /*Update links*/ wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1703.06907 Sim-to-Real Transfer of Robotic Control with Dynamics Randomization] || Josh Tobin et al. || simulated data can be used to train robotic control policies that transfer well to the real world using dynamics randomization, bridging the gap between simulation and real-world data. |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | 2021 || [https://arxiv.org/abs/2103.11624 Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding] || Krishna D. Kamath et al. || Focuses on predicting diverse future trajectories, crucial for creating realistic scenarios in robotics simulations. |- | 2021 || [https://arxiv.org/abs/1912.06680 Augmenting Reinforcement Learning with Human Videos] || Alex X. Lee et al. || Explores the use of human demonstration videos to improve the performance of reinforcement learning agents, which is highly relevant for augmenting datasets in robotics. |- | 2024 || [https://arxiv.org/abs/2402.05741 Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | 2024 || [https://arxiv.org/abs/2405.03520 Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | 2024 || [https://arxiv.org/abs/2403.09631 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | 2024 || [https://arxiv.org/abs/2403.09631 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | 2024 || [https://arxiv.org/abs/2402.02385 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | 2024 || [https://arxiv.org/abs/2402.06665 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | 2024 || [https://arxiv.org/abs/2306.06561 Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies . |- | 2024 || [https://arxiv.org/abs/2311.09064 Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |} 3b6a032eb352d41811b236b7a9f4182f578a943b 1602 1601 2024-06-28T06:35:16Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1703.06907 Sim-to-Real Transfer of Robotic Control with Dynamics Randomization] || Josh Tobin et al. || simulated data can be used to train robotic control policies that transfer well to the real world using dynamics randomization, bridging the gap between simulation and real-world data. |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | 2021 || [https://arxiv.org/abs/2103.11624 Diverse and Admissible Trajectory Forecasting through Multimodal Context Understanding] || Krishna D. Kamath et al. || Focuses on predicting diverse future trajectories, crucial for creating realistic scenarios in robotics simulations. |- | 2024 || [https://arxiv.org/abs/2402.05741 Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | 2024 || [https://arxiv.org/abs/2405.03520 Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | 2024 || [https://arxiv.org/abs/2403.09631 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | 2024 || [https://arxiv.org/abs/2403.09631 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | 2024 || [https://arxiv.org/abs/2402.02385 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | 2024 || [https://arxiv.org/abs/2402.06665 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | 2024 || [https://arxiv.org/abs/2306.06561 Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies . |- | 2024 || [https://arxiv.org/abs/2311.09064 Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |} dbdb3b157ac35d650b6d87fb22004a8e339627a0 1603 1602 2024-06-28T06:35:53Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1703.06907 Sim-to-Real Transfer of Robotic Control with Dynamics Randomization] || Josh Tobin et al. || simulated data can be used to train robotic control policies that transfer well to the real world using dynamics randomization, bridging the gap between simulation and real-world data. |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | 2024 || [https://arxiv.org/abs/2402.05741 Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | 2024 || [https://arxiv.org/abs/2405.03520 Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | 2024 || [https://arxiv.org/abs/2403.09631 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | 2024 || [https://arxiv.org/abs/2403.09631 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | 2024 || [https://arxiv.org/abs/2402.02385 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | 2024 || [https://arxiv.org/abs/2402.06665 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | 2024 || [https://arxiv.org/abs/2306.06561 Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies . |- | 2024 || [https://arxiv.org/abs/2311.09064 Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |} 464b4b4d7337479e2b8300824589a1e8a8e90643 1604 1603 2024-06-28T06:36:49Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |- | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | 2024 || [https://arxiv.org/abs/2402.05741 Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | 2024 || [https://arxiv.org/abs/2405.03520 Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | 2024 || [https://arxiv.org/abs/2403.09631 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | 2024 || [https://arxiv.org/abs/2403.09631 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | 2024 || [https://arxiv.org/abs/2402.02385 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | 2024 || [https://arxiv.org/abs/2402.06665 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | 2024 || [https://arxiv.org/abs/2306.06561 Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies . |- | 2024 || [https://arxiv.org/abs/2311.09064 Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |} 3307e5473ba5f1a25c431e4303d84c95c468ce88 1605 1604 2024-06-28T06:39:06Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | data-sort-value="2024-01-01" | 2024 || [https://arxiv.org/abs/2402.05741 Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | data-sort-value="2024-01-02" | 2024 || [https://arxiv.org/abs/2405.03520 Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | data-sort-value="2024-01-03" | 2024 || [https://arxiv.org/abs/2403.09631 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | data-sort-value="2024-01-04" | 2024 || [https://arxiv.org/abs/2403.09631 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | data-sort-value="2024-01-05" | 2024 || [https://arxiv.org/abs/2402.02385 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | data-sort-value="2024-01-06" | 2024 || [https://arxiv.org/abs/2402.06665 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | data-sort-value="2024-01-07" | 2024 || [https://arxiv.org/abs/2306.06561 Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies. |- | data-sort-value="2024-01-08" | 2024 || [https://arxiv.org/abs/2311.09064 Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |- | data-sort-value="2020-01-01" | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | data-sort-value="2018-01-01" | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | data-sort-value="2017-01-01" | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |} 923a5036ecb80d79222a108a5db7c530557f9312 1606 1605 2024-06-28T06:40:09Z Vrtnis 21 wikitext text/x-wiki World models leverage video data to create rich, synthetic datasets, enhancing the learning process for robotic systems. By generating diverse and realistic training scenarios, world models address the challenge of insufficient real-world data, enabling robots to acquire and refine skills more efficiently. {| class="wikitable sortable" ! Date !! Title !! Authors !! Summary |- | data-sort-value="2024-01-01" | 2024 || [https://arxiv.org/abs/2402.05741 Real-world Robot Applications of Foundation Models: A Review] || K Kawaharazuka, T Matsushima et al. || overview of the practical application of foundation models in real-world robotics, including the integration of specific components within existing robot systems. |- | data-sort-value="2024-01-02" | 2024 || [https://arxiv.org/abs/2405.03520 Is SORA a World Simulator? A Comprehensive Survey on General World Models and Beyond] || Z Zhu, X Wang, W Zhao, C Min, N Deng, M Dou et al. || surveys the applications of world models in various fields, including robotics, and discusses the potential of the SORA framework as a world simulator. |- | data-sort-value="2024-01-03" | 2024 || [https://arxiv.org/html/2401.04334v1 Large Language Models for Robotics: Opportunities, Challenges, and Perspectives] || J Wang, Z Wu, Y Li, H Jiang, P Shu, E Shi, H Hu et al. || perspectives of using large language models in robotics, focusing on model transparency, robustness, safety, and real-world applicability. |- | data-sort-value="2024-01-04" | 2024 || [https://arxiv.org/abs/2403.09631 3D-VLA: A 3D Vision-Language-Action Generative World Model] || H Zhen, X Qiu, P Chen, J Yang, X Yan, Y Du et al. || Presents 3D-VLA, a generative world model that combines vision, language, and action to guide robot control and achieve goal objectives. |- | data-sort-value="2024-01-05" | 2024 || [https://arxiv.org/abs/2402.02385 A Survey on Robotics with Foundation Models: Toward Embodied AI] || Z Xu, K Wu, J Wen, J Li, N Liu, Z Che, J Tang || integration of foundation models in robotics, addressing safety and interpretation challenges in real-world scenarios, particularly in densely populated environments. |- | data-sort-value="2024-01-06" | 2024 || [https://arxiv.org/abs/2402.06665 The Essential Role of Causality in Foundation World Models for Embodied AI] || T Gupta, W Gong, C Ma, N Pawlowski, A Hilmkil et al. || importance of causality in foundation world models for embodied AI, predicting that these models will simplify the introduction of new robots into everyday life. |- | data-sort-value="2024-01-07" | 2024 || [https://arxiv.org/abs/2306.06561 Learning World Models with Identifiable Factorization] || Y Liu, B Huang, Z Zhu, H Tian et al. || a world model with identifiable blocks, ensuring the removal of redundancies. |- | data-sort-value="2024-01-08" | 2024 || [https://arxiv.org/abs/2311.09064 Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models] || Y Kim, G Singh, J Park et al. || systematic generalization in vision models and world models. |- | data-sort-value="2020-01-01" | 2020 || [https://arxiv.org/abs/2003.08934 NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis] || Ben Mildenhall et al. || high-fidelity views of complex 3D scenes, instrumental in creating synthetic data for robotics, and relevant for generating diverse visual environments for training robots. |- | data-sort-value="2018-01-01" | 2018 || [https://arxiv.org/abs/1803.10122 World Models] || David Ha and Jürgen Schmidhuber || agent builds a compact model of the world and uses it to plan and dream, improving its performance in real environments. This aligns well with the interest in universal simulators. |- | data-sort-value="2017-01-01" | 2017 || [https://arxiv.org/abs/1612.07828 Learning from Simulated and Unsupervised Images through Adversarial Training] || Ashish Shrivastava et al. || technique that refines simulated images to make them more realistic using adversarial training, enhancing the quality of synthetic data for training robotics models. |} 8b1896b5f0c52e4cb39d6d21662b56343598c35c 0 345 1600 2024-06-27T21:22:57Z Vrtnis 21 Vrtnis moved page [[World Models In Robotics Learning]] to [[World Models]] wikitext text/x-wiki #REDIRECT [[World Models]] 4fd59cdb8f5af78295fbb22177b329ad9e7f6123 0 32 1607 1058 2024-06-28T18:57:53Z EtcetFelix 38 /* Isaac Sim Integration with Isaac Gym */ wikitext text/x-wiki Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with example [[applications]]. Typically you need a simulator, training framework and machine learning method to train end to end behaviors. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. For a much more comprehensive overview see [https://simulately.wiki/docs/ Simulately]. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===MuJoCo=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ===Bullet=== Bullet is a physics engine supporting real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning ==Simulators== ===[[Isaac Sim]]=== Isaac Sim is NVIDIA’s simulation platform for robotics development. It’s part of their Isaac Robotics platform and uses advanced graphics and AI to create realistic simulations. ==== Isaac Sim Features ==== * '''Advanced Physics Simulation''': Includes PhysX and Flex for detailed simulations of physical interactions like rigid bodies, soft bodies, and fluids. * '''Photorealistic Rendering''': Uses NVIDIA RTX technology to make environments and objects look incredibly realistic, which is great for tasks that need vision-based learning. * '''Scalability''': Can simulate multiple robots and environments at the same time, thanks to GPU acceleration, making it handle complex simulations efficiently. * '''Interoperability''': Works with machine learning frameworks like TensorFlow and PyTorch and supports ROS, so you can easily move from simulation to real-world deployment. * '''Customizable Environments''': Lets you create and customize simulation environments, including importing 3D models and designing different terrains. * '''Real-Time Feedback''': Provides real-time monitoring and analytics, giving you insights on how tasks are performing and resource usage. ==== Isaac Sim Applications ==== * '''Robotics Research''': Used in academia and industry to develop and test new algorithms for robot perception, control, and planning. * '''Autonomous Navigation''': Helps simulate and test navigation algorithms for mobile robots and drones, improving path planning and obstacle avoidance. * '''Manipulation Tasks''': Supports developing robotic skills like object grasping and assembly tasks, making robots more dexterous and precise. * '''Industrial Automation''': Helps companies design and validate automation solutions for manufacturing and logistics, boosting efficiency and cutting down on downtime. * '''Education and Training''': A great educational tool that offers hands-on experience in robotics and AI without the risks and costs of physical experiments. === Isaac Sim Integration with Isaac Gym === Note: Isaac Gym is now deprecated. NVIDIA now forwards users to their improved toolkit, Isaac Lab, also built on top of Isaac Sim.<br> Isaac Sim works alongside Isaac Gym, NVIDIA’s tool for large-scale training with reinforcement learning. While Isaac Sim focuses on detailed simulations, Isaac Gym is great for efficient training. Together, they offer a comprehensive solution for developing and improving robotics applications. ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== =Training frameworks= Popular training frameworks are listed here with example applications. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] 01bc1ab39239c051fdbe88fa19931392ee4b481a 1608 1607 2024-06-28T19:03:16Z EtcetFelix 38 /* Training frameworks */ wikitext text/x-wiki Learning algorithms allow to train humanoids to perform different skills such as manipulation or locomotion. Below is an overview of general approaches to training machine learning models for humanoid robots with example [[applications]]. Typically you need a simulator, training framework and machine learning method to train end to end behaviors. == Physics engines == Physics engines are software libraries designed to simulate physical systems in a virtual environment. They are crucial in a variety of fields such as video games, animation, robotics, and engineering simulations. These engines handle the mathematics involved in simulating physical processes like motion, collisions, and fluid dynamics. For a much more comprehensive overview see [https://simulately.wiki/docs/ Simulately]. ===PhysX=== PhysX is a physics engine by NVIDIA used primarily for video games and real-time simulations. It supports rigid body dynamics, cloth simulation, and particle effects, enhancing realism and interactivity in 3D environments. ===MuJoCo=== MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research in robotics and biomechanics. It's known for its speed, accuracy, and ease of use, making it popular for simulating complex systems with robotics and articulated structures. ===Bullet=== Bullet is a physics engine supporting real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning ==Simulators== ===[[Isaac Sim]]=== Isaac Sim is NVIDIA’s simulation platform for robotics development. It’s part of their Isaac Robotics platform and uses advanced graphics and AI to create realistic simulations. ==== Isaac Sim Features ==== * '''Advanced Physics Simulation''': Includes PhysX and Flex for detailed simulations of physical interactions like rigid bodies, soft bodies, and fluids. * '''Photorealistic Rendering''': Uses NVIDIA RTX technology to make environments and objects look incredibly realistic, which is great for tasks that need vision-based learning. * '''Scalability''': Can simulate multiple robots and environments at the same time, thanks to GPU acceleration, making it handle complex simulations efficiently. * '''Interoperability''': Works with machine learning frameworks like TensorFlow and PyTorch and supports ROS, so you can easily move from simulation to real-world deployment. * '''Customizable Environments''': Lets you create and customize simulation environments, including importing 3D models and designing different terrains. * '''Real-Time Feedback''': Provides real-time monitoring and analytics, giving you insights on how tasks are performing and resource usage. ==== Isaac Sim Applications ==== * '''Robotics Research''': Used in academia and industry to develop and test new algorithms for robot perception, control, and planning. * '''Autonomous Navigation''': Helps simulate and test navigation algorithms for mobile robots and drones, improving path planning and obstacle avoidance. * '''Manipulation Tasks''': Supports developing robotic skills like object grasping and assembly tasks, making robots more dexterous and precise. * '''Industrial Automation''': Helps companies design and validate automation solutions for manufacturing and logistics, boosting efficiency and cutting down on downtime. * '''Education and Training''': A great educational tool that offers hands-on experience in robotics and AI without the risks and costs of physical experiments. === Isaac Sim Integration with Isaac Gym === Note: Isaac Gym is now deprecated. NVIDIA now forwards users to their improved toolkit, Isaac Lab, also built on top of Isaac Sim.<br> Isaac Sim works alongside Isaac Gym, NVIDIA’s tool for large-scale training with reinforcement learning. While Isaac Sim focuses on detailed simulations, Isaac Gym is great for efficient training. Together, they offer a comprehensive solution for developing and improving robotics applications. ===[https://github.com/haosulab/ManiSkill ManiSkill]=== ===[[VSim]]=== =Training frameworks= Popular training frameworks are listed here with example applications. ===[https://isaac-sim.github.io/IsaacLab/index.html Isaac Lab]=== Isaac Lab is NVIDIA's modular framework for robot learning that aims to simplify common workflows in robotics research, part of the Isaac SDK. It is the successor to Isaac Gym. ===[https://github.com/NVIDIA-Omniverse/IsaacGymEnvs Isaac Gym]=== Note: Isaac Gym is now deprecated in favor of Isaac Lab. Isaac Gym is NVIDIA's robotics simulation tool, part of the Isaac SDK. It leverages GPU acceleration to enable the simulation of thousands of robot bodies simultaneously, making it highly efficient for training machine learning models in robotics. It's designed to streamline robotics applications, focusing on reinforcement learning in a virtual environment. ===[https://gymnasium.farama.org/ Gymnasium]=== Gymnasium is an open-source toolkit for developing and comparing reinforcement learning algorithms. Originally developed by OpenAI as "Gym," it provides a standardized set of environments (like Atari games, robotic simulations, etc.) to test and benchmark AI algorithms. It's widely used in the AI research community to foster innovation and replication in RL studies. ===[[Applications]]=== Over the last decade several advancements have been made in learning locomotion and manipulation skills with simulations. See non-comprehensive list here. == Training methods == ===[[Imitation learning]]=== Imitation Learning is a technique where models learn to perform tasks by mimicking expert behaviors. This approach is often used when defining explicit reward functions is challenging. It accelerates learning by using pre-collected datasets of expert demonstrations, reducing the need for trial-and-error in initial learning phases. ===[[Reinforcement Learning]]=== Reinforcement Learning involves agents learning to make decisions by interacting with an environment to maximize cumulative rewards. It's foundational in fields where sequential decision-making is crucial, like gaming, autonomous vehicles, and robotics. RL uses methods like Q-learning and policy gradient to iteratively improve agent performance based on feedback from the environment. [[Category: Software]] eb58fb71cbd97c3413752b079537abb384e1915c 0 34 1609 1158 2024-06-28T19:17:22Z EtcetFelix 38 /* Resources */ wikitext text/x-wiki This guide is incomplete and a work in progress; you can help by expanding it! == Reinforcement Learning (RL) == Reinforcement Learning (RL) is a machine learning approach where an agent learns to perform tasks by interacting with an environment. It involves the agent receiving rewards or penalties based on its actions and using this feedback to improve its performance over time. RL is particularly useful in robotics for training robots to perform complex tasks autonomously. Here's how RL is applied in robotics, using simulation environments like Isaac Sim and MuJoCo: == RL in Robotics == === Practical Applications of RL === ==== Task Automation ==== * Robots can be trained to perform repetitive or dangerous tasks autonomously, such as assembly line work, welding, or hazardous material handling. * RL enables robots to adapt to new tasks without extensive reprogramming, making them versatile for various industrial applications. ==== Navigation and Manipulation ==== * RL is used to train robots for navigating complex environments and manipulating objects with precision, which is crucial for tasks like warehouse logistics, domestic chores, and medical surgeries. === Simulation Environments === ==== Isaac Sim ==== * Isaac Sim provides a highly realistic and interactive environment where robots can be trained safely and efficiently. * The simulated environment includes physics, sensors, and other elements that mimic real-world conditions, enabling the transfer of learned behaviors to physical robots. ==== MuJoCo ==== * MuJoCo (Multi-Joint dynamics with Contact) is a physics engine designed for research and development in robotics, machine learning, and biomechanics. * It offers fast and accurate simulations, which are essential for training RL agents in tasks involving complex dynamics and contact-rich interactions. == Training algorithms == * [https://en.wikipedia.org/wiki/Advantage_Actor_Critic A2C] * [https://en.wikipedia.org/wiki/Proximal_policy_optimization PPO] * [https://spinningup.openai.com/en/latest/algorithms/sac.html SAC] == Resources == * [https://mandi-zhao.gitbook.io/deeprl-notes Mandy Zhao's Reinforcement Learning Notes] * [https://cs224r.stanford.edu/slides/cs224r-actor-critic-split.pdf Stanford CS224R Actor Critic Slides] * [https://farama.org/Announcing-The-Farama-Foundation RL Environment API by Farama Explanation] [[Category: Software]] [[Category: Reinforcement Learning]] dbd22d009cabcb5a3f4b220187b7358e2fef9754 0 1 1616 1589 2024-07-01T14:41:19Z Ben 2 /* List of Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/k3y9ymM9 Discord] b1a6a036f3445b279a66edebb322f6c5f3fdb040 1633 1616 2024-07-06T20:26:26Z Ben 2 /* List of Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/k3y9ymM9 Discord] f2c45f3ce63254c657df1bd2cb2ceeb4d4098018 1644 1633 2024-07-06T21:00:09Z Ben 2 /* Discord community */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] f20183aac92439a1e0b1f141e786023389e4d43c 0 294 1627 1594 2024-07-05T16:53:25Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1809263616958374286 2024.07.05] |- | [https://x.com/kscalelabs/status/1804184936574030284 2024.06.21] |- | [https://x.com/kscalelabs/status/1801749382167204086 2024.06.14] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] 08e2f9e423b533cd542b47f0e73289956df33f60 0 367 1634 2024-07-06T20:28:19Z Ben 2 Created page with "PaXini is a Chinese startup building a wheeled humanoid robot called [[Tora]]. They are based in Shenzhen. {{infobox company | name = PaXini Technology | country = China | we..." wikitext text/x-wiki PaXini is a Chinese startup building a wheeled humanoid robot called [[Tora]]. They are based in Shenzhen. {{infobox company | name = PaXini Technology | country = China | website_link = https://paxini.com/ | robots = [[Tora]] }} 148bcb4680512f2e2dbc7345e6bce0edce972337 0 368 1635 2024-07-06T20:29:05Z Ben 2 Created page with "Tora is a humanoid robot from the company [[PaXini]] {{infobox robot | name = Tora | organization = [[PaXini]] | cost = USD 20,000 | purchase_link = https://www.alibaba.com/p..." wikitext text/x-wiki Tora is a humanoid robot from the company [[PaXini]] {{infobox robot | name = Tora | organization = [[PaXini]] | cost = USD 20,000 | purchase_link = https://www.alibaba.com/product-detail/PAXINI-Tactile-Humanoid-Robot-With-Tactile_1601047975093.html }} 68384017e41070b2c4dc9bccaba9c584694f8a02 0 24 1636 253 2024-07-06T20:30:05Z Ben 2 wikitext text/x-wiki [[File:Cassie.jpg|right|200px|thumb]] Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility]]. {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = USD 250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} == Development == On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. [[Category: Robots]] 1e7db9d6a525aeb38ccddf6d846ed5b4c3dc028c 1650 1636 2024-07-06T21:39:28Z Ben 2 wikitext text/x-wiki [[File:Cassie.jpg|right|200px|thumb]] Cassie is a bipedal robot designed by Oregon State University and licensed and built by [[Agility]]. <youtube>https://www.youtube.com/watch?v=64hKiuJ31a4</youtube> {{infobox robot | name = Cassie | organization = [[Agility]] | video_link = https://www.youtube.com/watch?v=64hKiuJ31a4 | cost = USD 250,000 | height = 115 cm | weight = 31 kg | speed = >4 m/s (8.95 mph) | battery_life = 5 hours | battery_capacity = 1 kWh | dof = 10 (5 per leg) | number_made = ~12 | status = Retired }} == Development == On February 9, 2017 Cassie was unveiled at an event at Oregon State University featuring a live demo. A YouTube video was also posted to both the OSU YouTube page and Agility Robotics YouTube page. On September 5th, 2017 University of Michigan received the first Cassie which they named "Cassie Blue." They would later receive "Cassie Yellow." ==Firsts and World Record== Oregon State University's DRAIL lab used a reinforcement learned model to have Cassie run a 100M dash in 24.73 seconds. The Guinness World Record required a standing start and finish, so Cassie averaged a speed of over 4 m/s. For comparison, at the time the Guinness World Record for fastest running humanoid is ASIMO with a speed of 2.5 m/s which has to be done indoors on a special leveled floor. The run was done outside at the Whyte Track and Field Center on Oregon State's campus. The record was announced in September of 2022, but the actual run was 6 months prior. Cassie and the OSU DRAIL lab have an entire page dedicated to them in the 2024 Guinness World Records book. Footage of the run can be seen [https://www.youtube.com/watch?v=DdojWYOK0Nc here]. [[Category: Robots]] dc48d81db3806995deca7e583283a69c677204af 0 13 1637 885 2024-07-06T20:33:30Z Ben 2 wikitext text/x-wiki This is a course taught by Russ Tedrake at MIT. It is available [https://underactuated.csail.mit.edu/ here]. [[Category:Courses]] c5fce9af9e8669ceeaafd20ef1882dee30344564 0 55 1638 494 2024-07-06T20:52:50Z Ben 2 wikitext text/x-wiki [[File:1x Neo.jpg|thumb]] {{infobox robot | name = NEO | organization = [[1X]] | height = 165 cm | weight = 30 kg | video_link = https://www.youtube.com/watch?v=ikg7xGxvFTs | speed = 4 km/hr | carry_capacity = 20 kg | runtime = 2-4 hrs }} NEO is a bipedal humanoid robot developed by [[1X]]. It is designed to look and move like a human, featuring a head, torso, arms, and legs. NEO can perform a wide range of tasks, excelling in industrial sectors like security, logistics, manufacturing, operating machinery, and handling complex tasks. It is also envisioned to provide valuable home assistance and perform chores like cleaning or organizing. NEO's soft, tendon based design is meant to have very low inertia, intended to work in close proximity to humans. It will weigh 30 kilograms, with a 20 kilogram carrying capacity. 1X hopes for Neo to be "an all-purpose android assistant to your daily life." <youtube>https://www.youtube.com/watch?v=ikg7xGxvFTs</youtube> [[Category:Robots]] == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. 740c1ae8087d01657ba0cb6e66b9c8652d97879e 1639 1638 2024-07-06T20:53:10Z Ben 2 wikitext text/x-wiki [[File:1x Neo.jpg|thumb]] <youtube>https://www.youtube.com/watch?v=ikg7xGxvFTs</youtube> NEO is a bipedal humanoid robot developed by [[1X]]. It is designed to look and move like a human, featuring a head, torso, arms, and legs. NEO can perform a wide range of tasks, excelling in industrial sectors like security, logistics, manufacturing, operating machinery, and handling complex tasks. It is also envisioned to provide valuable home assistance and perform chores like cleaning or organizing. {{infobox robot | name = NEO | organization = [[1X]] | height = 165 cm | weight = 30 kg | video_link = https://www.youtube.com/watch?v=ikg7xGxvFTs | speed = 4 km/hr | carry_capacity = 20 kg | runtime = 2-4 hrs }} NEO's soft, tendon based design is meant to have very low inertia, intended to work in close proximity to humans. It will weigh 30 kilograms, with a 20 kilogram carrying capacity. 1X hopes for Neo to be "an all-purpose android assistant to your daily life." [[Category:Robots]] == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. 0b5e86ed46130dc85a95c95506702daed0384c7a 0 54 1640 451 2024-07-06T20:53:22Z Ben 2 wikitext text/x-wiki EVE is a versatile and agile humanoid robot developed by [[1X]]. It is equipped with cameras and sensors to perceive and interact with its surroundings. EVE’s mobility, dexterity, and balance allow it to navigate complex environments and manipulate objects effective. <youtube>https://www.youtube.com/watch?v=20GHG-R9eFI</youtube> {{infobox robot | name = EVE | organization = [[1X]] | height = 186 cm | weight = 86 kg | speed = 14.4 km/hr | carry_capacity = 15 kg | runtime = 6 hrs | video_link = https://www.youtube.com/watch?v=20GHG-R9eFI }} [[Category:Robots]] da8acccb10e80c932677af9a49d07480a3d4e254 0 286 1641 1297 2024-07-06T20:56:21Z Ben 2 wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} == Sensors == {| class="wikitable" ! Sensor type !! Number !! Notes |- | Cameras || 2 || Mounted in the eyes (see above), Pointgrey Dragonfly 2 (640x480) |- | Microphones || 2 || SoundMan High quality Stereo Omnidirectional microphone, -46 dB, 10V, 20....20 000 Hz +/- 3dB |- | Inertial sensors || 3+3 || Three axis gyroscopes + three axis accelerometers + three axis geomagnetic sensor based on BOSCH BNO055 chip, mounted in the head. (100Hz) |- | Joint sensors || For each large joint || Absolute magnetic encoder (12bit resolution @1kHz) at the joint, high-resolution incremental encoder at the motor side, hall-effect sensors for commutation (brushless motors only) |- | Joint sensors || For each small joint || Absolute magnetic encoder (except the fingers which use a custom hall-effect sensor), medium-resolution incremental encoder at the motor |- | Force/torque sensors || 6 || 6x6-axis force/torque sensors are mounted on the upper part of the arm and legs plus 2 additional sensors mounted closer to the ankle for higher precision ZMP estimation (100Hz) |- | Tactile sensors || More than 3000 (*) || Capacitive tactile sensors (8 bit resolution at 40Hz) are installed in the fingertips, palms, upper and fore-arms, chest and optionally at the legs (*). |} {| class="wikitable" |+ Capabilities of iCub ! Task !! Description |- | Crawling || Using visual guidance with an optic marker on the floor |- | Solving complex 3D mazes || Demonstrated ability to navigate and solve intricate 3D mazes |- | Archery || Shooting arrows with a bow and learning to hit the center of the target |- | Facial expressions || Capable of expressing emotions through facial expressions |- | Force control || Utilizing proximal force/torque sensors for precise force control |- | Grasping small objects || Able to grasp and manipulate small objects such as balls and plastic bottles |- | Collision avoidance || Avoids collisions within non-static environments and can also avoid self-collision |} == Links == * [https://www.iit.it/research/lines/icub IIT official website on iCub] * [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] a33e1758b65bfdc25d1b83f64bbd20cf364f564d 1642 1641 2024-07-06T20:57:14Z Ben 2 wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} == Sensors == {| class="wikitable" ! Sensor type !! Number !! Notes |- | Cameras || 2 || Mounted in the eyes (see above), Pointgrey Dragonfly 2 (640x480) |- | Microphones || 2 || SoundMan High quality Stereo Omnidirectional microphone, -46 dB, 10V, 20....20 000 Hz +/- 3dB |- | Inertial sensors || 3+3 || Three axis gyroscopes + three axis accelerometers + three axis geomagnetic sensor based on BOSCH BNO055 chip, mounted in the head. (100Hz) |- | Joint sensors || For each large joint || Absolute magnetic encoder (12bit resolution @1kHz) at the joint, high-resolution incremental encoder at the motor side, hall-effect sensors for commutation (brushless motors only) |- | Joint sensors || For each small joint || Absolute magnetic encoder (except the fingers which use a custom hall-effect sensor), medium-resolution incremental encoder at the motor |- | Force/torque sensors || 6 || 6x6-axis force/torque sensors are mounted on the upper part of the arm and legs plus 2 additional sensors mounted closer to the ankle for higher precision ZMP estimation (100Hz) |- | Tactile sensors || More than 3000 (*) || Capacitive tactile sensors (8 bit resolution at 40Hz) are installed in the fingertips, palms, upper and fore-arms, chest and optionally at the legs (*). |} {| class="wikitable" |+ Capabilities of iCub ! Task !! Description |- | Crawling || Using visual guidance with an optic marker on the floor |- | Solving complex 3D mazes || Demonstrated ability to navigate and solve intricate 3D mazes |- | Archery || Shooting arrows with a bow and learning to hit the center of the target |- | Facial expressions || Capable of expressing emotions through facial expressions |- | Force control || Utilizing proximal force/torque sensors for precise force control |- | Grasping small objects || Able to grasp and manipulate small objects such as balls and plastic bottles |- | Collision avoidance || Avoids collisions within non-static environments and can also avoid self-collision |} == Links == * [https://www.iit.it/research/lines/icub IIT official website on iCub] * [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] {{#related:Instituto Italiano}] [[Category:Robots]] [[Category:Humanoid Robots]] 77442c64b572e511bbd6fc3e36c5f731336ca297 1643 1642 2024-07-06T20:57:36Z Ben 2 /* Links */ wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} == Sensors == {| class="wikitable" ! Sensor type !! Number !! Notes |- | Cameras || 2 || Mounted in the eyes (see above), Pointgrey Dragonfly 2 (640x480) |- | Microphones || 2 || SoundMan High quality Stereo Omnidirectional microphone, -46 dB, 10V, 20....20 000 Hz +/- 3dB |- | Inertial sensors || 3+3 || Three axis gyroscopes + three axis accelerometers + three axis geomagnetic sensor based on BOSCH BNO055 chip, mounted in the head. (100Hz) |- | Joint sensors || For each large joint || Absolute magnetic encoder (12bit resolution @1kHz) at the joint, high-resolution incremental encoder at the motor side, hall-effect sensors for commutation (brushless motors only) |- | Joint sensors || For each small joint || Absolute magnetic encoder (except the fingers which use a custom hall-effect sensor), medium-resolution incremental encoder at the motor |- | Force/torque sensors || 6 || 6x6-axis force/torque sensors are mounted on the upper part of the arm and legs plus 2 additional sensors mounted closer to the ankle for higher precision ZMP estimation (100Hz) |- | Tactile sensors || More than 3000 (*) || Capacitive tactile sensors (8 bit resolution at 40Hz) are installed in the fingertips, palms, upper and fore-arms, chest and optionally at the legs (*). |} {| class="wikitable" |+ Capabilities of iCub ! Task !! Description |- | Crawling || Using visual guidance with an optic marker on the floor |- | Solving complex 3D mazes || Demonstrated ability to navigate and solve intricate 3D mazes |- | Archery || Shooting arrows with a bow and learning to hit the center of the target |- | Facial expressions || Capable of expressing emotions through facial expressions |- | Force control || Utilizing proximal force/torque sensors for precise force control |- | Grasping small objects || Able to grasp and manipulate small objects such as balls and plastic bottles |- | Collision avoidance || Avoids collisions within non-static environments and can also avoid self-collision |} == Links == * [https://www.iit.it/research/lines/icub IIT official website on iCub] * [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] a33e1758b65bfdc25d1b83f64bbd20cf364f564d 1660 1643 2024-07-06T21:44:13Z Ben 2 wikitext text/x-wiki The iCub is a research-grade humanoid robot for developing and testing embodied AI algorithms. The iCub Project integrates results from various [[Instituto Italiano]] Research Units. The iCub Project is a key initiative for IIT, aiming to transfer robotics technologies to industrial applications. {{infobox robot | name = iCub | organization = [[Instituto Italiano]] | height = 104 cm (3 ft 5 in) | weight = 22 kg (48.5 lbs) | video_link = https://www.youtube.com/watch?v=ErgfgF0uwUo | cost = Approximately €250,000 }} <youtube>https://www.youtube.com/watch?v=ErgfgF0uwUo</youtube> == General Specifications == The number of degrees of freedom is as follows: {| class="wikitable" ! Component !! # of degrees of freedom !! Notes |- | Eyes || 3 || Independent vergence and common tilt |- | Head || 3 || The neck has three degrees of freedom to tilt, swing, and pan |- | Chest || 3 || The torso can also tilt, swing, and pan |- | Arms || 7 (each) || The shoulder has 3 DoF, 1 in the elbow, and three in the wrist |- | Hands || 9 || The hand has 19 joints coupled in various combinations: the thumb, index, and middle finger are independent (coupled distal phalanxes), the ring and little finger are coupled. The thumb can additionally rotate over the palm. |- | Legs || 6 (each) || 6 DoF are sufficient to walk. |} == Sensors == {| class="wikitable" ! Sensor type !! Number !! Notes |- | Cameras || 2 || Mounted in the eyes (see above), Pointgrey Dragonfly 2 (640x480) |- | Microphones || 2 || SoundMan High quality Stereo Omnidirectional microphone, -46 dB, 10V, 20....20 000 Hz +/- 3dB |- | Inertial sensors || 3+3 || Three axis gyroscopes + three axis accelerometers + three axis geomagnetic sensor based on BOSCH BNO055 chip, mounted in the head. (100Hz) |- | Joint sensors || For each large joint || Absolute magnetic encoder (12bit resolution @1kHz) at the joint, high-resolution incremental encoder at the motor side, hall-effect sensors for commutation (brushless motors only) |- | Joint sensors || For each small joint || Absolute magnetic encoder (except the fingers which use a custom hall-effect sensor), medium-resolution incremental encoder at the motor |- | Force/torque sensors || 6 || 6x6-axis force/torque sensors are mounted on the upper part of the arm and legs plus 2 additional sensors mounted closer to the ankle for higher precision ZMP estimation (100Hz) |- | Tactile sensors || More than 3000 (*) || Capacitive tactile sensors (8 bit resolution at 40Hz) are installed in the fingertips, palms, upper and fore-arms, chest and optionally at the legs (*). |} {| class="wikitable" |+ Capabilities of iCub ! Task !! Description |- | Crawling || Using visual guidance with an optic marker on the floor |- | Solving complex 3D mazes || Demonstrated ability to navigate and solve intricate 3D mazes |- | Archery || Shooting arrows with a bow and learning to hit the center of the target |- | Facial expressions || Capable of expressing emotions through facial expressions |- | Force control || Utilizing proximal force/torque sensors for precise force control |- | Grasping small objects || Able to grasp and manipulate small objects such as balls and plastic bottles |- | Collision avoidance || Avoids collisions within non-static environments and can also avoid self-collision |} == Links == * [https://www.iit.it/research/lines/icub IIT official website on iCub] * [https://www.youtube.com/watch?v=znF1-S9JmzI Presentation of iCub by IIT] [[Category:Robots]] [[Category:Humanoid Robots]] 5a33ae5294017f6aa74d69fc79eb1f96c0c75191 1 63 1645 266 2024-07-06T21:12:45Z Ben 2 wikitext text/x-wiki You should install a YouTube plugin so we can embed robot demo videos! --[[User:Modeless|Modeless]] ([[User talk:Modeless|talk]]) 22:26, 24 April 2024 (UTC) : Added :) --[[User:Ben|Ben]] ([[User talk:Ben|talk]]) 21:12, 6 July 2024 (UTC) 1608584b4fe83acb4b666140f57fc963ddb06b70 0 123 1646 467 2024-07-06T21:38:39Z Ben 2 wikitext text/x-wiki RAISE-A1 is the first-generation general embodied intelligent robot developed by [[AGIBOT]]. The robot showcases industry-leading capabilities in bipedal walking intelligence and human-machine interaction, and is designed for use in various fields such as flexible manufacturing, interactive services, education and research, specialized avatars, warehousing logistics, and robotic household assistants. <youtube>https://www.youtube.com/watch?v=PIYJtZmzs70</youtube> {{infobox robot | name = RAISE-A1 | organization = [[AGIBOT]] | height = 175 cm | weight = 55 kg | single_arm_payload = 5 kg | runtime = 5 Hrs | walk_speed = 7 km/h | video_link = https://www.youtube.com/watch?v=PIYJtZmzs70 | cost = }} [[Category:Robots]] cd53f6a7c27f42f61db92278b4a958269743efb1 0 117 1647 466 2024-07-06T21:38:59Z Ben 2 wikitext text/x-wiki NASA’s Valkyrie, also known as R5, is a robust, rugged, and entirely electric humanoid robot. It was designed and built by the Johnson Space Center (JSC) Engineering Directorate to compete in the 2013 DARPA Robotics Challenge (DRC) Trials. <youtube>https://www.youtube.com/watch?v=LaYlQYHXJio</youtube> {{infobox robot | name = Valkyrie | organization = [[NASA]] | height = 190 cm | weight = 125 kg | video_link = https://www.youtube.com/watch?v=LaYlQYHXJio | cost = }} [[Category:Robots]] d1bea38266c75dbb8b9394640a35a0d2d9aeaa22 0 129 1648 444 2024-07-06T21:39:08Z Ben 2 wikitext text/x-wiki Draco is a high-performance bipedal platform developed by [[Apptronik]]. It’s their first biped robot, designed with a focus on speed and power. The system has 10 Degrees of Freedom (DOFs), allowing for a wide range of movements and tasks. One of the key features of Draco is its liquid cooling system, which helps maintain optimal performance during operation. <youtube>https://www.youtube.com/watch?v=g9xt2zdSOo8</youtube> {{infobox robot | name = Draco | organization = [[Apptronik]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=g9xt2zdSOo8 | cost = }} [[Category:Robots]] 7ac132a836fe11b924e75182c532779c5903f761 0 131 1649 448 2024-07-06T21:39:18Z Ben 2 wikitext text/x-wiki The humanoid robot Apollo is a creation of [[Apptronik]], a company known for its advanced humanoid robots. Apollo is a practical bipedal platform that’s designed to perform useful tasks. It’s equipped with two NVIDIA Jetson units and has been trained in the Isaac platform’s simulation environment. <youtube>https://www.youtube.com/watch?v=3CdwPGC9nyk&t=6s</youtube> {{infobox robot | name = Apollo | organization = [[Apptronik]] | height = 172.7 cm | weight = 73 kg | two_hand_payload = 25 | video_link = https://www.youtube.com/watch?v=3CdwPGC9nyk&t=6s | cost = }} [[Category:Robots]] 86fb2f0143b306925bdc4e22cb83e606f5f47092 0 128 1651 518 2024-07-06T21:39:43Z Ben 2 wikitext text/x-wiki [[File:Agility Robotics Digit.jpg|thumb]] Digit is a humanoid robot developed by [[Agility]], designed to navigate our world and perform tasks like navigation, obstacle avoidance, and manipulation. It's equipped with a torso full of sensors and a pair of arms, and is considered the most advanced Mobile Manipulation Robot (MMR) on the market, capable of performing repetitive tasks in production environments without requiring significant infrastructure changes. <youtube>https://www.youtube.com/watch?v=NgYo-Wd0E_U</youtube> {{infobox robot | name = Digit | organization = [[Agility]] | height = 175.3 cm | weight = 65 kg | two_hand_payload = 15.88 | runtime = | walk_speed = | video_link = https://www.youtube.com/watch?v=NgYo-Wd0E_U | cost = }} Digit is notably designed to be bipedal, but not necessarily humanoid/anthropomorphic, with ostrich-like reverse jointed legs. This is a side effect of Agility's design goals, to maximize efficiency and robustness of legged locomotion. == References == * Ackerman, Evan (2024). "Humanoid Robots are Getting to Work.", ''IEEE Spectrum''. [[Category:Robots]] e979751cc450d410d341c192c89fa7dec0ae2b3c 0 151 1652 619 2024-07-06T21:40:21Z Ben 2 wikitext text/x-wiki The Astribot S1 boasts an array of features that sets it apart from its predecessors. In its design, it intended to operate without the need for human guidance amidst its tasks, further emphasizing its autonomous ability<ref>https://www.geeky-gadgets.com/astribot-s1-ai-humanoid-robot/</ref>. The robot is known to assist in a variety of housework, including preparing drinks and doing other household chores such as ironing and folding<ref>https://www.msn.com/en-us/news/technology/astribot-ai-powered-humanoid-torso-can-prepare-drinks-help-with-housework/ar-AA1nP0ah</ref>. {{infobox robot | name = Astribot | organization = AstriBot Corporation | video_link = | cost = | height = | weight = | speed = 10 m/s | lift_force = 10 kg | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = 7 (per arm) | status = }} <youtube>https://www.youtube.com/watch?v=AePEcHIIk9s</youtube> Key specifications of the Astribot S1 include a top speed of 10 meters per second, which is faster than the average adult man, a payload capacity of 10 kg per arm and a total of 7 degrees of freedom per arm. These arms exhibit a range of motion rivaling that of a human limb<ref>https://digialps.com/meet-the-incredibly-fast-astribot-s1-the-humanoid-robot-that-learns-at-1x-speed-without-any-help/</ref>. == Capabilities == Astribot S1 is described as an 'all-purpose' home robot. While the full extent of its capabilities are still being unveiled, it is said to have the potential for tasks like folding clothes, cooking meals, and cleaning rooms<ref>https://businessnewsforkids.substack.com/p/85-a-super-cool-home-robot</ref>. In addition to these home tasks, the Astribot S1 is also adept at performing intricate tasks such as opening bottles and pouring drinks, which clearly illustrates the robot's well-designed dexterity<ref>https://www.msn.com/en-us/news/technology/astribot-ai-powered-humanoid-torso-can-prepare-drinks-help-with-housework/ar-AA1nP0ah</ref>. == References == <references /> [[Category:Robots]] f932c3d3a04d0e06caf9cad659279645041aa6ed 0 150 1653 609 2024-07-06T21:41:05Z Ben 2 wikitext text/x-wiki Sharing the name of the Chinese space station for some reason, this robot is claimed to be the "first" running electric humanoid. Unitree H1 already runs faster though. <youtube>https://www.youtube.com/watch?v=DeQpz5ycInE</youtube> == Links == * [https://www.maginative.com/article/meet-tiangong-chinas-full-size-electric-running-humanoid-robot/ Announcement article] 3f94a22ca7235e87e001a2eb53b172a1d172c703 0 109 1654 600 2024-07-06T21:41:33Z Ben 2 wikitext text/x-wiki Nadia is a humanoid robot developed by the Institute for Human & Machine Cognition (IHMC) in collaboration with Boardwalk Robotics<ref>[http://robots.ihmc.us/nadia Nadia Humanoid — IHMC Robotics Lab]</ref>. This advanced humanoid robot, named Nadia, is designed with a high power-to-weight ratio and a large range of motion, characteristics which provide it with exceptional mobility<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. Though specific information regarding height, weight, and payload capacities have not been explicitly stated, Nadia reportedly has one of the highest ranges of motion across its 29 joints of any humanoid robot globally<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. {{infobox robot | name = Nadia | organization = [[IHMC, Boardwalk Robotics]] | height = | weight = | single_hand_payload = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=uTmUfOc7r_s | cost = }} <youtube>https://www.youtube.com/watch?v=uTmUfOc7r_s</youtube> == Design and Capabilities == The Nadia humanoid robot's design encompasses a high power-to-weight ratio, contributing to its significant mobility potential. It stands out due to its extensive range of motion, facilitated by its architecture of 29 joints<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. These design features enable Nadia to adapt to and function within urban environments, aligning with the project's goal of facilitating semi-autonomous behaviors. Built with the same intelligence that powers the IHMC's DRC-Atlas robot, Nadia boasts real-time perception, compliant locomotion, autonomous footstep placement, and dexterous VR-teleoperated manipulation<ref>[https://www.boardwalkrobotics.com/Nadia.html Nadia Humanoid Robot - Boardwalk Robotics]</ref>. == Research and Development == The development of Nadia is a collaborative project by the IHMC Robotics Lab and Boardwalk Robotics. The research team aims to produce a next-generation humanoid, capable of executing more perilous tasks while retaining high mobility<ref>[https://www.bbc.com/news/world-us-canada-67722014 A VR-controlled robot that throws boxing punches - BBC]</ref>. This development project positions Nadia as one of the most mobile ground robots designed in-house at IHMC in nearly a decade<ref>[https://www.ihmc.us/news20221005/ Video shows progress of IHMC humanoid robot Nadia]</ref>. == References == <references /> [[Category:Robots]] aecf195316df94461500efd764a83537b447c923 0 134 1655 462 2024-07-06T21:42:04Z Ben 2 wikitext text/x-wiki HD Atlas is a highly dynamic humanoid robot developed by [[Boston Dynamics]]. It’s designed for real-world applications and is capable of demonstrating advanced athletics and agility. {{infobox robot | name = HD Atlas | organization = [[Boston Dynamics]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=-9EM5_VFlt8 | cost = }} <youtube>https://www.youtube.com/watch?v=-9EM5_VFlt8</youtube> [[Category:Robots]] 9a6352c71df769baf1b47d9c1311cd81339bde64 0 81 1656 330 2024-07-06T21:42:16Z Ben 2 wikitext text/x-wiki Atlas is a humanoid robot from [[Boston Dynamics]]. {{infobox robot | name = Atlas | organization = [[Boston Dynamics]] | height = 150 cm | weight = 89 kg | video_link = https://www.youtube.com/watch?v=29ECwExc-_M | single_hand_payload | two_hand_payload | cost = }} <youtube>https://www.youtube.com/watch?v=29ECwExc-_M</youtube> [[Category:Robots]] a815cf9a82a31291e4cec67e1de3d4a8b28d239d 0 77 1657 460 2024-07-06T21:42:37Z Ben 2 wikitext text/x-wiki The XR4, also known as one of the Seven Fairies named Xiaozi, is a humanoid bipedal robot developed by [[Dataa Robotics]]. XR4 is made from lightweight high-strength carbon fiber composite material with over 60 intelligent flexible joints. {{infobox robot | name = XR4 | organization = [[DATAA Robotics]] | height = 168 cm | weight = 65 kg | video_link = https://www.youtube.com/watch?v=DUyUZcH5uUU | single_hand_payload | two_hand_payload | cost = }} <youtube>https://www.youtube.com/watch?v=DUyUZcH5uUU</youtube> [[Category:Robots]] ebdbae7d2c1f52b93282d41e3f687fc77a8d55e6 0 75 1658 574 2024-07-06T21:42:52Z Ben 2 wikitext text/x-wiki The Wukong-IV is an adult-size humanoid robot designed and built by the research team at [[Deep Robotics]]. It stands 1.4 meters tall and weighs 45 kg<ref>https://pdfs.semanticscholar.org/f4d3/80f8e0fe39906f21f5270ffd2bf7bae74039.pdf</ref>. This bionic humanoid robot is actuated by 21 electric motor joints. It has 6 degrees of freedom (DoF) on each leg and 4 DoFs on each arm<ref>https://www.mdpi.com/2072-666X/13/10/1688</ref>. {{infobox robot | name = Wukong-IV | organization = [[Deep Robotics]] | height = 140 cm | weight = 45 kg | single_hand_payload = | two_hand_payload = | cost = | video_link = https://www.youtube.com/watch?v=fbk4fYc6U14 | dof = 21 | number_made = | lift_force = | battery_life = | battery_capacity = | purchase_link = | status = | speed = }} <youtube>https://www.youtube.com/watch?v=fbk4fYc6U14</youtube> == References == <references /> [[Category:Robots]] 0f391d5c3658b6366b7b2e7a37198bd69535db99 0 103 1659 409 2024-07-06T21:43:08Z Ben 2 wikitext text/x-wiki MagicBot is a humanoid robot from [[MagicLab, DREAME]]. {{infobox robot | name = MagicBot | organization = [[MagicLab, DREAME]] | height = 178 cm | weight = 56 kg | single_hand_payload | two_hand_payload | video_link = https://www.youtube.com/watch?v=NTPmiDrHv4E&t=2s | cost }} <youtube>https://www.youtube.com/watch?v=NTPmiDrHv4E&t=2s</youtube> [[Category:Robots]] 898e7c7e1a2374fc1132c645e34ad451c7c42f44 0 94 1661 397 2024-07-06T21:44:28Z Ben 2 wikitext text/x-wiki Kaleido is a humanoid robot from [[Kawasaki Robotics]]. {{infobox robot | name = Kaleido | organization = [[Kawasaki Robotics]] | height = 179 cm | weight = 83 kg | two_hand_payload = 60 | video_link = https://www.youtube.com/watch?v=_h66xSbIEdU }} <youtube>https://www.youtube.com/watch?v=_h66xSbIEdU</youtube> [[Category:Robots]] 913fba144fa1aace4e3520635b0f0c8c0a96bc58 0 96 1662 399 2024-07-06T21:44:41Z Ben 2 wikitext text/x-wiki Friends is a humanoid robot from [[Kawasaki Robotics]]. {{infobox robot | name = Friends | organization = [[Kawasaki Robotics]] | height = 168 cm | weight = 54 kg | two_hand_payload = 10 | video_link = https://www.youtube.com/watch?v=dz4YLbgbVvc }} <youtube>https://www.youtube.com/watch?v=dz4YLbgbVvc</youtube> [[Category:Robots]] ebb5b61a346601caf191af85d9cc4c69e60816bc 0 92 1663 395 2024-07-06T21:45:06Z Ben 2 wikitext text/x-wiki K1 is a humanoid robot from [[Kepler]]. {{infobox robot | name = K1 | organization = [[Kepler]] | height = 178 cm | weight = 85 kg | video_link = https://www.youtube.com/watch?v=A5vshTgDbKE }} <youtube>https://www.youtube.com/watch?v=A5vshTgDbKE</youtube> [[Category:Robots]] 141efeba859c3efb1623feb514a8839555bc67ab 0 2 1664 1459 2024-07-06T21:45:25Z Ben 2 wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb|Stompy standing up]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] * [[Gripper History]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions === Conventions === The images below show our pin convention for the CAN bus when using various connectors. <gallery> Kscale db9 can bus convention.jpg Kscale phoenix can bus convention.jpg </gallery> = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. = Artwork = Here's some art of Stompy! <gallery> Stompy 1.png Stompy 2.png Stompy 3.png Stompy 4.png </gallery> [[Category:Robots]] [[Category:Open Source]] [[Category:K-Scale]] 70fe138ddcd37045d7a3e6bd2586109620a79b71 0 107 1665 413 2024-07-06T21:45:47Z Ben 2 wikitext text/x-wiki Mona is a humanoid robot from [[Kind Humanoid]]. {{infobox robot | name = Mona | organization = [[Kind Humanoid]] | height = | weight = | single_hand_payload | two_hand_payload = | video_link = https://www.youtube.com/watch?v=LsKOpooGK4Q }} <youtube>https://www.youtube.com/watch?v=LsKOpooGK4Q</youtube> [[Category:Robots]] 49716b38951bd5e434cdcdc762250bec3f0b6b20 0 215 1666 932 2024-07-06T21:46:29Z Ben 2 wikitext text/x-wiki Open-source humanoid robot from USC.<ref>https://github.com/DRCL-USC/Hector_Simulation</ref><ref>https://laser-robotics.com/hector-v2/</ref> <youtube>https://www.youtube.com/watch?v=W4f0641Kcpg</youtube> === References === <references/> 1c3b0da779df3d7f393efc04425313cc8148326c 0 318 1667 1458 2024-07-06T21:47:11Z Ben 2 wikitext text/x-wiki Reachy is a humanoid robot developed by [[Pollen Robotics]]. {{infobox robot | name = Reachy | organization = [[Pollen Robotics]] | video_link = https://www.youtube.com/watch?v=oZxHkp4-DnM }} <youtube>https://www.youtube.com/watch?v=oZxHkp4-DnM</youtube> [[Category:Robots]] 890f330121ebc30e69e232404b147248c3a1c7cc 0 83 1668 333 2024-07-06T21:47:21Z Ben 2 wikitext text/x-wiki HUBO is a humanoid robot from [[Rainbow Robotics]]. {{infobox robot | name = HUBO | organization = [[Rainbow Robotics]] | height = 170 cm | weight = 80 kg | video_link = https://www.youtube.com/watch?v=r2pKEVTddy4 | single_hand_payload | two_hand_payload | cost = USD 320,000 }} <youtube>https://www.youtube.com/watch?v=r2pKEVTddy4</youtube> [[Category:Robots]] b739f736e8aed14e4a05d8b544c9860ab5fd6a3b 1669 1668 2024-07-06T21:47:59Z Ben 2 wikitext text/x-wiki HUBO is a humanoid robot from [[Rainbow Robotics]]. {{infobox robot | name = HUBO | organization = [[Rainbow Robotics]] | height = 170 cm | weight = 80 kg | video_link = https://www.youtube.com/watch?v=r2pKEVTddy4 | single_hand_payload | two_hand_payload | cost = USD 320,000 }} [[Category:Robots]] 2ef8e6ee8ee0469684e4f490089a879a8fbd6535 0 132 1670 458 2024-07-06T21:48:12Z Ben 2 wikitext text/x-wiki XBot is a humanoid robot developed by [[Robot Era]], a startup incubated by Tsinghua University. The company has open-sourced its reinforcement learning framework, Humanoid-Gym, which was used to train the XBot and has proven successful in sim-to-real policy transfer. {{infobox robot | name = Xbot | organization = [[Robot Era]] | height = 122 cm | weight = 38 kg | two_hand_payload = 25 | video_link = https://www.youtube.com/watch?v=4tiVkZBw188 }} <youtube>https://www.youtube.com/watch?v=4tiVkZBw188</youtube> [[Category:Robots]] fd288868e509d5fd61a469daddcac49c3d442bd1 0 53 1671 535 2024-07-06T21:48:23Z Ben 2 wikitext text/x-wiki Phoenix is a sophisticated humanoid robot developed by [[Sanctuary AI]], a prominent company known for its advancements in robotics and artificial intelligence. {{infobox robot | name = Phoenix | organization = Sanctuary AI | video_link = https://youtube.com/watch?v=FH3zbUSMAAU | height = 5 ft 7 in (170 cm) | weight = 70 kg (155 lbs) | two_hand_payload = 25 kg }} <youtube>https://youtube.com/watch?v=FH3zbUSMAAU</youtube> === Development and Capabilities === Phoenix, introduced on May 16, 2024, represents the seventh generation of humanoid robots aimed at performing general-purpose tasks in various industries, including but not limited to service, healthcare, and logistics. This robot was designed to mimic human dexterity and mobility, allowing it to operate in environments built for human functionality. === Major Features === * '''Height and Weight''': Phoenix stands at a height of 170 cm and weighs approximately 70 kilograms, which is within the range of an average adult human. This anthropomorphic design facilitates easier integration into human-centric environments. * '''Two-Hand Payload''': The robot has a two-hand payload capacity of 25 kilograms, making it capable of handling substantial weight, which is essential for tasks involving lifting and carrying objects. === Public Release === Sanctuary AI publicly unveiled Phoenix on May 16, 2024, through a comprehensive announcement that highlighted the robot's potential applications and its contribution to advancing human-robot collaboration. The detailed introduction and capabilities were featured in a press release and a demonstration video, which showcased Phoenix performing a variety of tasks.<ref>Sanctuary AI News Release, ''Sanctuary AI Unveils Phoenix, a Humanoid General-Purpose Robot Designed for Work'', [https://sanctuary.ai/resources/news/sanctuary-ai-unveils-phoenix-a-humanoid-general-purpose-robot-designed-for-work/]</ref> [[File:Main-image-phoenix-annoucement.jpg|none|500px|Phoenix Gen 7|thumb]] == References == <references /> [[Category:Robots]] 12143c135a9c0358c115ff6bcce2bf9008883b14 0 243 1672 1030 2024-07-06T21:48:31Z Ben 2 wikitext text/x-wiki Pepper is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. Pepper is designed to interact with humans and is used in various customer service and retail environments. {{infobox robot | name = Pepper | organization = [[Softbank Robotics]] | height = 121 cm (4 ft) | weight = 28 kg (62 lbs) | video_link = https://www.youtube.com/watch?v=kr05reBxVRs | cost = Approximately $1,800 }} <youtube>https://www.youtube.com/watch?v=kr05reBxVRs</youtube> Pepper was introduced by SoftBank Robotics in June 2014. It is designed to understand and respond to human emotions, making it suitable for roles in customer service, retail, and healthcare. == References == [https://www.softbankrobotics.com/ SoftBank Robotics official website] [https://www.youtube.com/watch?v=2GhUd0OJdJw Presentation of Pepper by SoftBank Robotics] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:SoftBank Robotics]] 94ee68610b5af1acf49c917c0c3cb4799d506d1e 0 244 1673 1031 2024-07-06T21:48:40Z Ben 2 wikitext text/x-wiki NAO is a humanoid robot developed by [[Softbank Robotics]], a division of SoftBank Group. NAO is widely used in education, research, and healthcare for its advanced interactive capabilities. {{infobox robot | name = NAO | organization = [[Softbank Robotics]] | height = 58 cm (1 ft 11 in) | weight = 5.4 kg (11.9 lbs) | video_link = https://www.youtube.com/watch?v=nNbj2G3GmAo | cost = Approximately $8,000 }} <youtube>https://www.youtube.com/watch?v=nNbj2G3GmAo</youtube> NAO was first introduced in 2006 by Aldebaran Robotics, which was later acquired by SoftBank Robotics. NAO has undergone several upgrades, becoming one of the most popular robots used for educational and research purposes. == References == [https://www.softbankrobotics.com/ SoftBank Robotics official website] [https://www.youtube.com/watch?v=nNbj2G3GmAo Presentation of NAO by SoftBank Robotics] [[Category:Robots]] [[Category:Humanoid Robots]] [[Category:SoftBank Robotics]] 6c56dacb5c487715ed4db2fe381041e2042b5a04 0 114 1674 424 2024-07-06T21:49:06Z Ben 2 wikitext text/x-wiki Rocky is [[SuperDroid Robots]] most versatile robot platform which can move on any terrain, climb stairs, and step over obstacles. We are planning to open-source the humanoid once it is publicly walking and interacting with physical objects. {{infobox robot | name = Rocky | organization = [[SuperDroid Robots]] | video_link = https://twitter.com/stevenuecke/status/1707899032973033690 | cost = $75,000 | height = 64 in | weight = 120 pounds | speed = | lift_force = 150 lbs | battery_life = 8 hours | battery_capacity = 3,600 Wh | purchase_link = https://www.superdroidrobots.com/humanoid-biped-robot/ | number_made = 1 | dof = 22 | status = Finishing sim-2-real for motors }} <youtube>https://www.youtube.com/watch?v=MvAS4AsMvCI</youtube> [[Category:Robots]] 0de569b82f8f5221974b39eb0feecffd7c24909b 0 79 1675 461 2024-07-06T21:50:03Z Ben 2 wikitext text/x-wiki ZEUS2Q is a humanoid robot developed by [[System Technology Works]]. It’s a stand-alone system that harnesses the power of Edge AI computing, enabling it to perform localized AI tasks like communication, facial, and object recognition on the edge. {{infobox robot | name = ZEUS2Q | organization = [[System Technology Works]] | height = 127 cm | weight = 13.61 kg | video_link = https://www.youtube.com/watch?v=eR2HMykMITY }} <youtube>https://www.youtube.com/watch?v=eR2HMykMITY</youtube> [[Category:Robots]] ea79f0125e895cd77531bd3d15a8ba8c6f2d656f 0 22 1676 622 2024-07-06T21:50:18Z Ben 2 wikitext text/x-wiki [[File:Optimus Tesla (1).jpg|right|200px|thumb]] Optimus is a humanoid robot developed by [[Tesla]], an American electric vehicle and clean energy company. Also known as Tesla Bot, Optimus is a key component of Tesla's expansion into automation and artificial intelligence technologies. {{infobox robot | name = Optimus | organization = [[Tesla]] | height = 5 ft 8 in (173 cm) | weight = 58 kg | video_link = https://www.youtube.com/watch?v=cpraXaw7dyc | cost = Unknown, rumored $20k }} <youtube>https://www.youtube.com/watch?v=cpraXaw7dyc</youtube> == Development == Tesla initiated the development of the Optimus robot in 2021, with the goal of creating a multipurpose utility robot capable of performing unsafe, repetitive, or boring tasks primarily intended for a factory setting. Tesla's CEO, Elon Musk, outlined that Optimus could potentially transition into performing tasks in domestic environments in the future. == Design == The robot stands at a height of 5 feet 8 inches and weighs approximately 58 kilograms. Its design focusses on replacing human labor in hazardous environments, incorporating advanced sensors and algorithms to navigate complex workspaces safely. == Features == The features of Optimus are built around its capability to handle tools, carry out tasks requiring fine motor skills, and interact safely with human environments. The robot is equipped with Tesla's proprietary Full Self-Driving (FSD) computer, allowing it to understand and navigate real-world scenarios effectively. == Impact == Optimus has significant potential implications for labor markets, particularly in industries reliant on manual labor. Its development also sparks discussions on ethics and the future role of robotics in society. == References == * [https://www.tesla.com Tesla official website] * [https://www.youtube.com/watch?v=cpraXaw7dyc Presentation of Optimus by Tesla] [[Category:Robots]] fd381a2e80b1a65407c54b67427cf078de013405 0 137 1677 601 2024-07-06T21:50:28Z Ben 2 wikitext text/x-wiki I wasn't able to find the specific height, weight, or cost of the Punyo robot. However, I was able to find general information about the robot, which employed whole-body manipulation, using its arms and chest, to assist with everyday tasks<ref>https://newatlas.com/robotics/toyota-punyo-humanoid-soft-robot/</ref><ref>https://punyo.tech/</ref>. The robot platform is constructed with compliant materials and tactile mechanisms that increase its ability to safely handle objects, especially large and heavy ones<ref>https://medium.com/toyotaresearch/meet-punyo-tris-soft-robot-for-whole-body-manipulation-research-949c934ac3d8</ref><ref>https://spectrum.ieee.org/humanoid-robot-tri-punyo</ref>. Here is the updated information for the Punyo robot: {{infobox robot | name = Punyo | organization = [[Toyota Research Institute]] | height = | weight = | two_hand_payload = | video_link = https://www.youtube.com/watch?v=FY-MD4gteeE }} <youtube>https://www.youtube.com/watch?v=FY-MD4gteeE</youtube> == Article == Punyo is a soft robot developed by the [[Toyota Research Institute]] (TRI) to revolutionize whole-body manipulation research. Unlike traditional robots that primarily use hands for manipulation, Punyo employs its arms and chest. The robot is designed to help with everyday tasks, such as lifting heavy objects or closing a drawer<ref>https://newatlas.com/robotics/toyota-punyo-humanoid-soft-robot/</ref><ref>https://punyo.tech/</ref>. == Description == Punyo's platform designs with compliant materials and employs tactile mechanisms that increase its ability to handle objects effectively and safely, especially large and heavy ones<ref>https://medium.com/toyotaresearch/meet-punyo-tris-soft-robot-for-whole-body-manipulation-research-949c934ac3d8</ref><ref>https://spectrum.ieee.org/humanoid-robot-tri-punyo</ref>. == References == <references /> [[Category:Robots]] f53dab9a8c7600806de6e7fe32bd5b5161d51743 0 135 1678 475 2024-07-06T21:50:41Z Ben 2 wikitext text/x-wiki The T-HR3 is the third generation humanoid robot unveiled by [[Toyota Research Institute]]. It’s designed to explore new technologies for safely managing physical interactions between robots and their surroundings, and it features a remote maneuvering system that mirrors user movements to the robot. The T-HR3 can assist humans in various settings such as homes, medical facilities, construction sites, disaster-stricken areas, and even outer space. {{infobox robot | name = T-HR3 | organization = [[Toyota Research Institute]] | height = 150 cm | weight = 75 kg | two_hand_payload = | video_link = https://www.youtube.com/watch?v=5dPY7l7u_z0 }} <youtube>https://www.youtube.com/watch?v=5dPY7l7u_z0</youtube> [[Category:Robots]] af386d373c3c159b3929e9834b04b6fe8e9ab226 0 71 1679 454 2024-07-06T21:50:50Z Ben 2 wikitext text/x-wiki Walker X is a highly advanced AI humanoid robot developed by [[UBTECH]]. It incorporates six cutting-edge AI technologies, including upgraded vision-based navigation and hand-eye coordination, enabling it to move smoothly and quickly, and engage in precise and safe interactions. it is equipped with 41 high-performance servo joints, a 160° face surrounding 4.6K HD dual flexible curved screen, and a 4-dimensional light language system. {{infobox robot | name = Walker X | organization = [[UBTech]] | height = 130 cm | weight = 63 kg | single_hand_payload = 1.5 | two_hand_payload = 3 | cost = USD 960,000 | video_link = https://www.youtube.com/watch?v=4ZL3LgdKNbw }} <youtube>https://www.youtube.com/watch?v=4ZL3LgdKNbw</youtube> [[Category:Robots]] b1151429ed4406064a83362e9a372e0fd5a9601a 0 74 1680 456 2024-07-06T21:51:01Z Ben 2 wikitext text/x-wiki Walker S by [[UBTech]] is a highly advanced humanoid robot designed to serve in household and office scenarios. It is equipped with 36 high-performance servo joints and a full range of sensory systems including force, vision, hearing, and spatial awareness, enabling smooth and fast walking and flexible, precise handling. {{infobox robot | name = Walker S | organization = [[UBTech]] | height = | weight = | video_link = https://www.youtube.com/watch?v=UCt7qPpTt-g | single_hand_payload = | two_hand_payload = | cost = }} <youtube>https://www.youtube.com/watch?v=UCt7qPpTt-g</youtube> [[Category:Robots]] ff95dcceac1127abc162585795db721c26df8643 0 3 1681 612 2024-07-06T21:51:17Z Ben 2 wikitext text/x-wiki '''Unitree H1''' is a full-size universal humanoid robot developed by the [[Unitree]], a company known for its innovative robotic designs. The H1 is celebrated for its superior power performance capabilities and advanced powertrain technologies. {{infobox robot | name = H1 | organization = [[Unitree]] | video_link = https://www.youtube.com/watch?v=83ShvgtyFAg | cost = USD 150,000 | height = 180 cm | weight = 47 kg | speed = >3.3 m/s | lift_force = | battery_life = | battery_capacity = 864 Wh | purchase_link = https://shop.unitree.com/products/unitree-h1 | number_made = | dof = | status = }} <youtube>https://www.youtube.com/watch?v=83ShvgtyFAg</youtube> == Specifications == The H1 robot stands approximately 180 cm tall and weighs around 47 kg, offering high mobility and physical capabilities. Some of the standout specifications of the H1 include: * Maximum speed: Exceeds 3.3 meters per second, a benchmark in robot mobility. * Weight: Approximately 47 kg. * Maximum joint torque: 360 N.m. * Battery capacity: 864 Wh, which is quickly replaceable, enhancing the robot's operational endurance. == Features == The H1 incorporates advanced technologies to achieve its high functionality: * Highly efficient powertrain for superior speed, power, and maneuverability. * Equipped with high-torque joint motors developed by [[Unitree]] itself. * 360° depth sensing capabilities combined with LIDAR and depth cameras for robust environmental perception. == Uses and Applications == While detailed use cases of Unitree H1 are not extensively documented, the robot's build and capabilities suggest its suitability in complicated tasks requiring human-like dexterity and strength which might include industrial applications, complex terrain navigation, and interactive tasks. [[Category:Robots]] 4b761c614db4f672f92bb904f8d61e017c3c91a0 0 233 1682 1022 2024-07-06T21:51:39Z Ben 2 wikitext text/x-wiki [[File:Unitree g1.png|thumb]] The G1 humanoid robot is an upcoming humanoid robot from [[Unitree]]. {{infobox robot | name = G1 | organization = [[Unitree]] | video_link = https://mp.weixin.qq.com/s/RGNVRazZqDn3y_Ijemc5Kw | cost = 16000 USD | height = 127 cm | weight = 35 kg | speed = | lift_force = | battery_life = | battery_capacity = 9000 mAh | purchase_link = | number_made = | dof = 23 | status = Preorders }} {{infobox robot | name = G1 Edu Standard | organization = [[Unitree]] | cost = 31900 USD | notes = Improved torque, warranty }} {{infobox robot | name = G1 Edu Plus | organization = [[Unitree]] | cost = 34900 USD | notes = Docking station }} {{infobox robot | name = G1 Edu Smart | organization = [[Unitree]] | cost = 43900 USD | notes = 3 waist DoFs instead of 1, more arm DoFs }} {{infobox robot | name = G1 Edu Ultimate | organization = [[Unitree]] | cost = 53900 USD | notes = Comes with force-controlled 3-finger dexterous hands }} <youtube>https://www.youtube.com/watch?v=GzX1qOIO1bE</youtube> 11e3ef845d8fe70dc02103f78f910840a1c90b84 0 115 1683 606 2024-07-06T21:51:58Z Ben 2 wikitext text/x-wiki While the search results have shown information about a robot named THEMIS, it seems to be about the THeMIS unmanned ground vehicle (UGV) developed by the company Milrem Robotics, as seen on the Milrem Robotics website<ref>[Milrem Robotics Defence](https://milremrobotics.com/defence/)</ref>. The search results do not provide additional information about the THEMIS humanoid robot from Westwood Robotics. Based on the initial information provided, we can update the infobox as follows (note that the cost field remains unfilled): {{infobox robot | name = THEMIS | organization = [[Westwood Robotics]] | video_link = https://www.youtube.com/watch?v=yt4mHwAl9cc | height = 142.2 cm | weight = 39 kg | cost = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} <youtube>https://www.youtube.com/watch?v=yt4mHwAl9cc</youtube> Further details about the robot's costs, speed, lift force, number made, degrees of freedom, battery life/capacity, and status are needed to complete the infobox. Until we are able to elaborate more on the THEMIS humanoid robot, we cannot finalize this article. == References == <references /> [[Category:Robots]] 5b1c744201b65b73242e0a658f92ea4ed67bd698 0 298 1684 1357 2024-07-06T21:52:10Z Ben 2 wikitext text/x-wiki The '''Syntro''' robot from [[WorkFar]] is designed for advanced warehouse automation. {{infobox robot | name = Syntro | organization = [[WorkFar]] | video_link = https://www.youtube.com/watch?v=suF7mEtLJvY | cost = | height = | weight = | speed = | lift_force = | battery_life = | battery_capacity = | purchase_link = | number_made = | dof = | status = }} <youtube>https://www.youtube.com/watch?v=suF7mEtLJvY</youtube> [[Category:Robots]] 517569dcd2fe44f6d116951259ee63aa30d1b7ae 0 126 1685 623 2024-07-06T21:52:17Z Ben 2 wikitext text/x-wiki CyberOne is a humanoid robot developed by the Chinese consumer electronics giant, Xiaomi. Unveiled in 2022 at a company event in Beijing by the founder, chairman, and CEO, Lei Jun, it is the newest member of Xiaomi's Cyber series, joining previously launched quadruped robots like CyberDog and CyberDog 2<ref>https://robotsguide.com/robots/cyberone</ref>. {{infobox robot | name = CyberOne | organization = [[Xiaomi]] | height = 177 cm | weight = 52 kg | single_arm_payload = 1.5 | runtime = | walk_speed = 3.6 km/h | video_link = https://www.youtube.com/watch?v=yBmatGQ0giY | cost = 600,000 - 700,000 yuan (est.) }} <youtube>https://www.youtube.com/watch?v=yBmatGQ0giY</youtube> == Specifications == This bipedal humanoid robot has a height of 177 cm and weight of 52 kg, with an arm span of 168 cm<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. One of its distinct features is its ability to perceive 3D, recognize individuals, and respond to human emotions<ref>https://www.gadgets360.com/smart-home/news/xiaomi-cyberone-unveiled-specifications-humanoid-bionic-robot-oled-display-features-3249173</ref>. Furthermore, it boasts a top speed of 3.6 km/ph<ref>https://www.theverge.com/2022/8/16/23307808/xiaomi-cyberone-humanoid-robot-tesla-optimus-bot-specs-comparison</ref>. == Pricing == The cost of CyberOne, if ever produced and made available for purchase, is estimated to be around 600,000 to 700,000 yuan<ref>https://robbreport.com/gear/electronics/xiaomi-humanoid-robot-cyberone-1234738597/</ref>. [[Category:Robots]] == References == <references /> 19c136625b7759c54563be287f22db3fb4a19a65 0 111 1686 625 2024-07-06T21:52:25Z Ben 2 wikitext text/x-wiki The PX5 is a humanoid robot developed by Xpeng, unveiled for the first time during Xpeng Motors' Tech Day in 2023.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> The robot stands approximately 1.5 meters in height and is able to navigate different terrain and handle objects with precision, demonstrating remarkable stability.<ref>https://technode.com/2023/10/25/xpeng-tech-day-2023-first-mpv-self-driving-timeline-flying-cars-and-humanoid-robots/</ref> Constructed with a silver-white color scheme, the PX5 is also resistant to shock.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> {{infobox robot | name = PX5 | organization = [[Xpeng]] | height = 1.5 meters | weight = | video_link = https://www.youtube.com/watch?v=BNSZ8Fwcd20 | cost = }} <youtube>https://www.youtube.com/watch?v=BNSZ8Fwcd20</youtube> == Development == Xpeng Robotics, an ecosystem company of Xpeng, specializing in smart robots, revealed the PX5. The company, which was founded in 2016, innovates in areas like robot powertrain, locomotion control, robot autonomy, robot interaction, and artificial intelligence contributing to a shared mission of exploring future mobility solutions.<ref>https://www.pxing.com/en/about</ref> == Design and Capabilities == The PX5 bears a striking silver-white finish and exhibits resistance to shock. Its capability to navigate through different terrains and handle handheld objects, such as a pen, with exceptional stability has been highlighted in demonstrations.<ref>https://kr-asia.com/xpeng-motors-unveils-px5-humanoid-robot-underlining-its-vision-for-the-future</ref> <ref>https://technode.com/2023/10/25/xpeng-tech-day-2023-first-mpv-self-driving-timeline-flying-cars-and-humanoid-robots/</ref> == References == <references /> [[Category:Robots]] 11fe68e066b6fe89def83c89e1530418982723ba 0 117 1694 1647 2024-07-07T18:02:09Z Ben 2 small cleanup wikitext text/x-wiki NASA’s Valkyrie, also known as R5, is a robust, rugged, and entirely electric humanoid robot. It was designed and built by the Johnson Space Center (JSC) Engineering Directorate to compete in the 2013 DARPA Robotics Challenge (DRC) Trials. <youtube>https://www.youtube.com/watch?v=LaYlQYHXJio</youtube> {{infobox robot | name = Valkyrie | organization = [[NASA]] | height = 190 cm | weight = 125 kg | video_link = https://www.youtube.com/watch?v=LaYlQYHXJio }} [[Category:Robots]] 01955b8fdce745d59aedf87289e62f126ca7c459 6 376 1695 2024-07-08T15:38:21Z Ben 2 wikitext text/x-wiki Robot taking notes 907ccf64e98bfb5d3b7ead7d4b154f86e3319dd0 0 294 1696 1627 2024-07-08T15:38:43Z Ben 2 wikitext text/x-wiki [[File:Robot taking notes.png|thumb|Robot taking notes (from Midjourney)]] {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1809263616958374286 2024.07.05] |- | [https://x.com/kscalelabs/status/1804184936574030284 2024.06.21] |- | [https://x.com/kscalelabs/status/1801749382167204086 2024.06.14] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] ab8ec4a8d3c203cd44b334b55184bafbfb23b600 1720 1696 2024-07-12T16:52:25Z Ben 2 wikitext text/x-wiki [[File:Robot taking notes.png|thumb|Robot taking notes (from Midjourney)]] {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1811805432505336073 2024.07.12] |- | [https://x.com/kscalelabs/status/1809263616958374286 2024.07.05] |- | [https://x.com/kscalelabs/status/1804184936574030284 2024.06.21] |- | [https://x.com/kscalelabs/status/1801749382167204086 2024.06.14] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] 5e8581190123da2179befb77f1d63c54860b696c 0 1 1697 1644 2024-07-08T21:51:27Z Vrtnis 21 /*Correct Figure AI wikilink*/ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] 54ba8b2ec9adbe65cb51f7ebfd0c3b9c47c505f6 1705 1697 2024-07-11T23:28:56Z Vrtnis 21 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] ea0d4924a37d38f19e1e0b8aa9b7a5a7d2315cd0 0 377 1698 2024-07-09T20:56:11Z Vrtnis 21 Created page with "Beijing has launched a new innovation center dedicated to humanoid robots, aiming to boost technological advancements and industrial growth. Located in the Beijing Economic-Te..." wikitext text/x-wiki Beijing has launched a new innovation center dedicated to humanoid robots, aiming to boost technological advancements and industrial growth. Located in the Beijing Economic-Technological Development Area, this center will tackle key industry challenges and support the development of globally influential robot companies by 2025. <ref>https://english.beijing.gov.cn/latest/news/202311/t20231105_3295012.html</ref> e09c62ff5f44286885f46c035984e82e5ad2fb4f 0 378 1699 2024-07-09T20:58:38Z Vrtnis 21 Created page with "The Anthro™ is an anthropomorphic (humanoid) host for embodied synthetic intelligence (or AGI). <ref>https://anthrobotics.ca/the-anthro/</ref>" wikitext text/x-wiki The Anthro™ is an anthropomorphic (humanoid) host for embodied synthetic intelligence (or AGI). <ref>https://anthrobotics.ca/the-anthro/</ref> e793ccd0e1c3165e6e930d27ff8f13c88a58605e 0 379 1700 2024-07-09T21:23:51Z Vrtnis 21 Created page with "Apptronik’s latest upper body humanoid robot is designed to operate with and around humans. It has state-of-the-art actuation packed into a small form factor and can be put..." wikitext text/x-wiki Apptronik’s latest upper body humanoid robot is designed to operate with and around humans. It has state-of-the-art actuation packed into a small form factor and can be put on any mobility platform. <ref>https://apptronik.com/our-work</ref> 2bab8570a014a5938c7c61502d1eb2b89cae27a3 1701 1700 2024-07-09T21:24:20Z Vrtnis 21 wikitext text/x-wiki [[Apptronik]]’s latest upper body humanoid robot is designed to operate with and around humans. It has state-of-the-art actuation packed into a small form factor and can be put on any mobility platform. <ref>https://apptronik.com/our-work</ref> d354462b3738851cc00288b839cfbc508cba5608 0 380 1702 2024-07-09T21:30:56Z Vrtnis 21 Created page with " BR002 is a humanoid robot developed by Booster Robotics, designed to showcase advanced mobility and versatility. It can transition from a prone position to standing by foldin..." wikitext text/x-wiki BR002 is a humanoid robot developed by Booster Robotics, designed to showcase advanced mobility and versatility. It can transition from a prone position to standing by folding its legs backward, similar to the method used by Boston Dynamics' Atlas. BR002 features complex motion control algorithms for its humanoid leg configurations, allowing it to perform lateral and vertical splits, with joints capable of 360° rotation, offering a wide range of motion. <ref> https://equalocean.com/news/2024050820854</ref> 3acf27253c8f724f7aaed0c52f7603e97596524b 0 11 1703 219 2024-07-10T22:25:37Z Vrtnis 21 wikitext text/x-wiki Skild is a stealth foundation model startup started by two faculty members from Carnegie Mellon University. {{infobox company | name = Skild | country = United States | website_link = https://www.skild.ai/ }} === Articles === * [https://www.theinformation.com/articles/venture-fomo-hits-robotics-as-young-startup-gets-1-5-billion-valuation Venture FOMO Hits Robotics as Young Startup Gets $1.5 Billion Valuation] [[Category:Companies]] e641ae138131c1ceb3f744bbf0d448761fe0e65a 0 381 1704 2024-07-11T23:27:01Z Vrtnis 21 Created page with "'''Engineered Arts''' is a UK-based company specializing in the development of humanoid robots and advanced robotic systems. Known for their sophisticated and lifelike creatio..." wikitext text/x-wiki '''Engineered Arts''' is a UK-based company specializing in the development of humanoid robots and advanced robotic systems. Known for their sophisticated and lifelike creations, they integrate cutting-edge technology to produce robots used in entertainment, research, and public engagement.<ref>https://engineeredarts.co.uk/</ref> 617d02d020f684da74dd2faaa652f7266edeabbd 0 382 1706 2024-07-11T23:30:55Z Vrtnis 21 Created page with "'''Ameca''' is a humanoid robot developed by Engineered Arts, known for its advanced human-like expressions and interactions. Designed to serve as a platform for artificial in..." wikitext text/x-wiki '''Ameca''' is a humanoid robot developed by Engineered Arts, known for its advanced human-like expressions and interactions. Designed to serve as a platform for artificial intelligence and machine learning research, Ameca is equipped with state-of-the-art sensors and actuators to mimic human movements and emotions.<ref>https://engineeredarts.co.uk/robot/ameca/</ref> fc4b0d6c57bfac63c19d57684c6c9fa570cc3840 0 383 1707 2024-07-11T23:32:28Z Vrtnis 21 Created page with "'''Kuavo (Kuafu)''' is a humanoid robot jointly developed by Haier Robot and Leju Robot, and was first exhibited at AWE 2024. As China's first open-source Hongmeng humanoid ro..." wikitext text/x-wiki '''Kuavo (Kuafu)''' is a humanoid robot jointly developed by Haier Robot and Leju Robot, and was first exhibited at AWE 2024. As China's first open-source Hongmeng humanoid robot designed for family scenes, Kuavo is capable of jumping, multi-terrain walking, and performing tasks such as washing, watering flowers, and drying clothes. The development of Kuavo follows a strategic cooperation between Haier Home Robot and Leju Robot to integrate AI and robotics into smart home environments.<ref>http://www.datayuan.cn/article/20862.htm</ref> 2ad07dc797fe79c2e83651fcc5a34099c5569502 0 384 1708 2024-07-11T23:35:54Z Vrtnis 21 Created page with "'''Shadow-1''' is Hyperspawn's flagship open-source robot, designed to perform complex tasks with human-like precision and efficiency, particularly suited for exploration and..." wikitext text/x-wiki '''Shadow-1''' is Hyperspawn's flagship open-source robot, designed to perform complex tasks with human-like precision and efficiency, particularly suited for exploration and task execution in extreme conditions. This advanced bipedal robot integrates cutting-edge technology to navigate and operate in challenging environments, showcasing Hyperspawn's commitment to innovation in robotic performance and reliability.<ref>https://www.hyperspawn.co/bipeds</ref> ded51a30de533cd3d41ca40594331c3f2666699f 0 385 1709 2024-07-11T23:37:30Z Vrtnis 21 Created page with "'''NEURA Robotics''' is a German high-tech company founded in 2019 in Metzingen near Stuttgart, dedicated to enhancing collaborative robots with cognitive capabilities for saf..." wikitext text/x-wiki '''NEURA Robotics''' is a German high-tech company founded in 2019 in Metzingen near Stuttgart, dedicated to enhancing collaborative robots with cognitive capabilities for safer and more efficient human-robot interaction. With a team of over 170 members from 30+ countries, NEURA Robotics has developed innovations like MAiRA, the world’s first cognitive robot, and MiPA, a versatile robotic assistant.<ref>https://neura-robotics.com/company</ref> e65efc64d9bd1df338e047b89ce24d0c276b5277 0 386 1710 2024-07-11T23:38:47Z Vrtnis 21 Created page with "'''Kayra''' is an open-source, easy-to-modify, 3D-printable humanoid robot. Designed for accessibility and customization, Kayra aims to make advanced robotics more approachabl..." wikitext text/x-wiki '''Kayra''' is an open-source, easy-to-modify, 3D-printable humanoid robot. Designed for accessibility and customization, Kayra aims to make advanced robotics more approachable for enthusiasts and researchers.<ref>https://kayra.org</ref> 6551581e8abb54f61ffbe992235213124ae7f7b1 0 387 1711 2024-07-11T23:40:10Z Vrtnis 21 Created page with "'''Noetix Robotics''' focuses on the development and manufacturing of humanoid robots, integrating fields such as general artificial intelligence, robot bionics, and embodied..." wikitext text/x-wiki '''Noetix Robotics''' focuses on the development and manufacturing of humanoid robots, integrating fields such as general artificial intelligence, robot bionics, and embodied operating systems. Their work aims to advance the capabilities and applications of humanoid robotics.<ref>https://noetixrobotics.com/</ref> b2435a008e3124105c5b36349b9d16bbb38a44d8 0 388 1712 2024-07-11T23:42:00Z Vrtnis 21 Created page with "'''Dora''' is a humanoid robot developed by Noetix Robotics, a start-up technology company established in Beijing in September 2023. Dora, showcased at the World Artificial In..." wikitext text/x-wiki '''Dora''' is a humanoid robot developed by Noetix Robotics, a start-up technology company established in Beijing in September 2023. Dora, showcased at the World Artificial Intelligence Conference, is China's first lightweight, commercialized general-purpose humanoid robot. It is designed for various applications, including scientific research, education, exhibition display, and service companionship.<ref>http://www.azchinesenews1.com/static/content/XW/2024-07-05/1258803435970535978.html</ref> 628e5e3d0534800c04a54625cbbed5b418fd8046 0 389 1713 2024-07-11T23:43:58Z Vrtnis 21 Created page with "'''POINTBLANK''' is a company specializing in rapid prototyping, hardware development, open-source projects, and robotics. They are dedicated to advancing technology through i..." wikitext text/x-wiki '''POINTBLANK''' is a company specializing in rapid prototyping, hardware development, open-source projects, and robotics. They are dedicated to advancing technology through innovative solutions and accessible designs.<ref>https://www.pointblankllc.com/</ref> 96c5cabfc124baed7c12f0d13f5ceaf749ee5a79 0 390 1714 2024-07-11T23:45:05Z Vrtnis 21 Created page with "'''Dropbear''' is an advanced humanoid robot developed by Hyperspawn and Pointblank. Designed to operate in varied environments, Dropbear showcases agility, precision, and int..." wikitext text/x-wiki '''Dropbear''' is an advanced humanoid robot developed by Hyperspawn and Pointblank. Designed to operate in varied environments, Dropbear showcases agility, precision, and intelligence.<ref>https://github.com/Hyperspawn/Dropbear</ref> 58c50090335a51faafdcb299a6546bd1ff5f9732 0 391 1715 2024-07-11T23:46:40Z Vrtnis 21 Created page with "'''Robotera''' was established in August 2023 and focuses on developing universal humanoid robots for various fields. Their latest product, the XBot-L, is a full-size humanoid..." wikitext text/x-wiki '''Robotera''' was established in August 2023 and focuses on developing universal humanoid robots for various fields. Their latest product, the XBot-L, is a full-size humanoid robot capable of navigating complex terrains and performing advanced tasks with high flexibility and accuracy. Xingdong Jiyuan or Robotera has received multiple prestigious awards and significant media attention, showcasing its innovation and leadership in the humanoid robot industry.<ref>https://www.robotera.com/</ref> 1f6ee75421a4e81b1903e9543b89934cce33b926 0 392 1716 2024-07-11T23:48:39Z Vrtnis 21 Created page with "'''Starbot''' is a state-of-the-art humanoid robot developed by Professor Jianyu Chen's startup, RobotEra, in collaboration with his research team at Tsinghua University. Star..." wikitext text/x-wiki '''Starbot''' is a state-of-the-art humanoid robot developed by Professor Jianyu Chen's startup, RobotEra, in collaboration with his research team at Tsinghua University. Starbot boasts exceptional motion capabilities, including traversing snowy terrains, navigating staircases, and robust interference resistance, all controlled by a single neural network trained via end-to-end reinforcement learning.<ref>https://mobile.x.com/roboterax/status/1742443820384718930</ref> 01c09e7beaf845425721960d8d86031e57d73774 0 393 1717 2024-07-11T23:49:55Z Vrtnis 21 Created page with "'''Navigator α''' is a newly launched humanoid robot standing 1.5 meters tall and weighing 50 kilograms. It features advanced hardware, including a planetary reducer, lightwe..." wikitext text/x-wiki '''Navigator α''' is a newly launched humanoid robot standing 1.5 meters tall and weighing 50 kilograms. It features advanced hardware, including a planetary reducer, lightweight humanoid arms, and a dexterous hand with 15 finger joints and six active degrees of freedom. Navigator α integrates mechanism control, imitation learning, and reinforcement learning, utilizing large-scale AI models for significant advancements in hardware and algorithms.<ref>https://www.prnewswire.com/news-releases/navigator--humanoid-robot-supcon-integrates-ai-with-robotics-302107204.html</ref> 0777669abea3e020f2ac5e64436e5b17f3ce26c2 0 394 1718 2024-07-11T23:51:22Z Vrtnis 21 Created page with "'''Berkeley Control, Intelligent Systems, and Robotics (CIR)''' at EECS focuses on modeling systems and machines to ensure appropriate responses to inputs, utilizing optimizat..." wikitext text/x-wiki '''Berkeley Control, Intelligent Systems, and Robotics (CIR)''' at EECS focuses on modeling systems and machines to ensure appropriate responses to inputs, utilizing optimization and mathematical techniques. Research spans semiconductor process control, hybrid and networked control, nonlinear and learning control, and involves collaboration with Mechanical Engineering and Integrative Biology. The robotics research covers mobile autonomous systems, human augmentation, telepresence, and virtual reality, with an emphasis on image understanding and computer vision.<ref>https://www2.eecs.berkeley.edu/Research/Areas/CIR/</ref> d3bc9a2d3a600e59f2f079f835e17c5180549c9d 0 395 1719 2024-07-11T23:52:42Z Vrtnis 21 Created page with "The Robotics Lab at the University of Tehran focuses on advanced research in robotics, including autonomous systems, human-robot interaction, and artificial intelligence. It a..." wikitext text/x-wiki The Robotics Lab at the University of Tehran focuses on advanced research in robotics, including autonomous systems, human-robot interaction, and artificial intelligence. It aims to develop innovative robotic solutions for industrial and service applications.<ref>https://robotics.ut.ac.ir/</ref> b434d796e55df9577f472075a0d0dda2db75d34d 0 299 1721 1591 2024-07-14T01:18:50Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | 2024.07.18 | Wesley | Consistency Models |- | 2024.07.17 | Nathan | Flow Matching |- | 2024.07.17 | Dennis | Diffusion Models |- | 2024.07.16 | Isaac | Convolutional Neural Networks |- | 2024.07.15 | Allen | RNNs, LSTMs, GRUs |- | 2024.06.28 | Kenji | Principles of Power Electronics |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.26 | Ryan | Introduction to KiCAD |- | 2024.06.24 | Dennis | Live Streaming Protocols |- | 2024.06.21 | Nathan | Quantization |- | 2024.06.20 | Timothy | Diffusion |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] e783ead99b090b2d6b918a5adce4406f044e3abb 1722 1721 2024-07-14T01:19:06Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | 2024.07.19 | Wesley | Consistency Models |- | 2024.07.18 | Nathan | Flow Matching |- | 2024.07.17 | Dennis | Diffusion Models |- | 2024.07.16 | Isaac | Convolutional Neural Networks |- | 2024.07.15 | Allen | RNNs, LSTMs, GRUs |- | 2024.06.28 | Kenji | Principles of Power Electronics |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.26 | Ryan | Introduction to KiCAD |- | 2024.06.24 | Dennis | Live Streaming Protocols |- | 2024.06.21 | Nathan | Quantization |- | 2024.06.20 | Timothy | Diffusion |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 7cea6ba1ca6b7728f827a94cd0b8199b37a9f8fb 0 396 1723 2024-07-18T16:31:52Z Vrtnis 21 Created page with "= Expressive Whole-Body Control for Humanoid Robots = == Installation == <code> conda create -n humanoid python=3.8 conda activate humanoid cd pip3 install torch==1.10.0+cu11..." wikitext text/x-wiki = Expressive Whole-Body Control for Humanoid Robots = == Installation == <code> conda create -n humanoid python=3.8 conda activate humanoid cd pip3 install torch==1.10.0+cu113 torchvision==0.11.1+cu113 torchaudio==0.10.0+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html git clone git@github.com:chengxuxin/expressive_humanoid_covariant.git cd expressive_humanoid_covariant # Download the Isaac Gym binaries from https://developer.nvidia.com/isaac-gym cd isaacgym/python && pip install -e . cd ~/expressive_humanoid_covariant/rsl_rl && pip install -e . cd ~/expressive_humanoid_covariant/legged_gym && pip install -e . pip install "numpy<1.24" pydelatin wandb tqdm opencv-python ipdb pyfqmr flask dill gdown </code> Next install fbx. Follow the instructions here. == Prepare dataset == Download from here and extract the zip file to ASE/ase/poselib/data/cmu_fbx_all that contains all .fbx files. Generate .yaml file for the motions you want to use. <code> cd ASE/ase/poselib python parse_cmu_mocap_all.py </code> This step is not mandatory because the .yaml file is already generated. But if you want to add more motions, you can use this script to generate the .yaml file. == Import motions == <code> cd ASE/ase/poselib python fbx_importer_all.py </code> This will import all motions in CMU Mocap dataset into ASE/ase/poselib/data/npy. == Retarget motions == <code> cd ASE/ase/poselib mkdir pkl retarget_npy python retarget_motion_h1_all.py </code> This will retarget all motions in ASE/ase/poselib/data/npy to ASE/ase/poselib/data/retarget_npy. == Generate keybody positions == This step will require running simulation to extract more precise key body positions. <code> cd legged_gym/legged_gym/scripts python train.py debug --task h1_view --motion_name motions_debug.yaml --debug </code> Train for 1 iteration and kill the program to have a dummy model to load. <code> python play.py debug --task h1_view --motion_name motions_autogen_all.yaml </code> It is recommended to use motions_autogen_all.yaml at the first time, so that later if you have a subset it is not necessary to regenerate keybody positions. This will generate keybody positions to ASE/ase/poselib/data/retarget_npy. Set wandb asset: == Usage == To train a new policy: <code> python train.py xxx-xx-some_descriptions_of_run --device cuda:0 --entity WANDB_ENTITY </code> xxx-xx is usually an id like 000-01. motion_type and motion_name are defined in legged_gym/legged_gym/envs/h1/h1_mimic_config.py. They can be also given as arguments. Can set default WANDB_ENTITY in legged_gym/legged_gym/utils/helpers.py. To play a policy: <code> python play.py xxx-xx </code> No need to write the full experiment id. The parser will auto match runs with first 6 strings (xxx-xx). So better make sure you don't reuse xxx-xx. Delay is added after 8k iters. If you want to play after 8k, add --delay. To play with example pretrained models: <code> python play.py 060-40 --delay --motion_name motions_debug.yaml </code> Try to press + or - to see different motions. The motion name will be printed on terminal. motions_debug.yaml is a small subset of motions for debugging and contains some representative motions. Save models for deployment: <code> python save_jit.py --exptid xxx-xx </code> This will save the models in legged_gym/logs/parkour_new/xxx-xx/traced/. == Viewer Usage == Can be used in both IsaacGym and web viewer. * ALT + Mouse Left + Drag Mouse: move view. * [ ]: switch to next/prev robot. * Space: pause/unpause. * F: switch between free camera and following camera. === IsaacGym viewer specific === * +: next motion in yaml. * -: prev motion in yaml. * r: reset the motion to the beginning. * ]: camera focus on next env. * [: camera focus on prev env. == Arguments == * --exptid: string, can be xxx-xx-WHATEVER, xxx-xx is typically numbers only. WHATEVER is the description of the run. * --device: can be cuda:0, cpu, etc. * --delay: whether add delay or not. * --checkpoint: the specific checkpoint you want to load. If not specified load the latest one. * --resume: resume from another checkpoint, used together with --resumeid. * --seed: random seed. * --no_wandb: no wandb logging. * --entity: specify wandb entity * --web: used for playing on headless machines. It will forward a port with vscode and you can visualize seamlessly in vscode with your idle gpu or cpu. Live Preview vscode extension required, otherwise you can view it in any browser. * --motion_name: e.g. 07_04 or motions_all.yaml. If motions_all.yaml is used, motion_type should be yaml. * --motion_type: single or yaml * --fix_base: fix the base of the robot. For more arguments, refer legged_gym/utils/helpers.py. 5fbe025d1f9569567fed8d882fb371c5d9a75120 1724 1723 2024-07-18T16:35:13Z Vrtnis 21 /*Add Refs*/ wikitext text/x-wiki = Expressive Whole-Body Control for Humanoid Robots = == Installation == <code> conda create -n humanoid python=3.8 conda activate humanoid cd pip3 install torch==1.10.0+cu113 torchvision==0.11.1+cu113 torchaudio==0.10.0+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html git clone git@github.com:chengxuxin/expressive_humanoid_covariant.git cd expressive_humanoid_covariant # Download the Isaac Gym binaries from https://developer.nvidia.com/isaac-gym cd isaacgym/python && pip install -e . cd ~/expressive_humanoid_covariant/rsl_rl && pip install -e . cd ~/expressive_humanoid_covariant/legged_gym && pip install -e . pip install "numpy<1.24" pydelatin wandb tqdm opencv-python ipdb pyfqmr flask dill gdown </code> Next install fbx. Follow the instructions here. == Prepare dataset == Download from here and extract the zip file to ASE/ase/poselib/data/cmu_fbx_all that contains all .fbx files. Generate .yaml file for the motions you want to use. <code> cd ASE/ase/poselib python parse_cmu_mocap_all.py </code> This step is not mandatory because the .yaml file is already generated. But if you want to add more motions, you can use this script to generate the .yaml file. == Import motions == <code> cd ASE/ase/poselib python fbx_importer_all.py </code> This will import all motions in CMU Mocap dataset into ASE/ase/poselib/data/npy. == Retarget motions == <code> cd ASE/ase/poselib mkdir pkl retarget_npy python retarget_motion_h1_all.py </code> This will retarget all motions in ASE/ase/poselib/data/npy to ASE/ase/poselib/data/retarget_npy. == Generate keybody positions == This step will require running simulation to extract more precise key body positions. <code> cd legged_gym/legged_gym/scripts python train.py debug --task h1_view --motion_name motions_debug.yaml --debug </code> Train for 1 iteration and kill the program to have a dummy model to load. <code> python play.py debug --task h1_view --motion_name motions_autogen_all.yaml </code> It is recommended to use motions_autogen_all.yaml at the first time, so that later if you have a subset it is not necessary to regenerate keybody positions. This will generate keybody positions to ASE/ase/poselib/data/retarget_npy. Set wandb asset: == Usage == To train a new policy: <code> python train.py xxx-xx-some_descriptions_of_run --device cuda:0 --entity WANDB_ENTITY </code> xxx-xx is usually an id like 000-01. motion_type and motion_name are defined in legged_gym/legged_gym/envs/h1/h1_mimic_config.py. They can be also given as arguments. Can set default WANDB_ENTITY in legged_gym/legged_gym/utils/helpers.py. To play a policy: <code> python play.py xxx-xx </code> No need to write the full experiment id. The parser will auto match runs with first 6 strings (xxx-xx). So better make sure you don't reuse xxx-xx. Delay is added after 8k iters. If you want to play after 8k, add --delay. To play with example pretrained models: <code> python play.py 060-40 --delay --motion_name motions_debug.yaml </code> Try to press + or - to see different motions. The motion name will be printed on terminal. motions_debug.yaml is a small subset of motions for debugging and contains some representative motions. Save models for deployment: <code> python save_jit.py --exptid xxx-xx </code> This will save the models in legged_gym/logs/parkour_new/xxx-xx/traced/. == Viewer Usage == Can be used in both IsaacGym and web viewer. * ALT + Mouse Left + Drag Mouse: move view. * [ ]: switch to next/prev robot. * Space: pause/unpause. * F: switch between free camera and following camera. === IsaacGym viewer specific === * +: next motion in yaml. * -: prev motion in yaml. * r: reset the motion to the beginning. * ]: camera focus on next env. * [: camera focus on prev env. == Arguments == * --exptid: string, can be xxx-xx-WHATEVER, xxx-xx is typically numbers only. WHATEVER is the description of the run. * --device: can be cuda:0, cpu, etc. * --delay: whether add delay or not. * --checkpoint: the specific checkpoint you want to load. If not specified load the latest one. * --resume: resume from another checkpoint, used together with --resumeid. * --seed: random seed. * --no_wandb: no wandb logging. * --entity: specify wandb entity * --web: used for playing on headless machines. It will forward a port with vscode and you can visualize seamlessly in vscode with your idle gpu or cpu. Live Preview vscode extension required, otherwise you can view it in any browser. * --motion_name: e.g. 07_04 or motions_all.yaml. If motions_all.yaml is used, motion_type should be yaml. * --motion_type: single or yaml * --fix_base: fix the base of the robot. For more arguments, refer legged_gym/utils/helpers.py. == References == * [https://expressive-humanoid.github.io/ Expressive Whole-Body Control for Humanoid Robots] * [https://chengxuxin.github.io/ Xuxin Cheng] * [https://yandongji.github.io/ Yandong Ji] * [https://jeremycjm.github.io/ Junming Chen] * [https://www.episodeyang.com/ Ge Yang] * [https://xiaolonw.github.io/ Xiaolong Wang] * Publisher: UC San Diego * Year: 2024 * [https://expressive-humanoid.github.io/ Website] * [https://arxiv.org/abs/2402.16796 arXiv] 5d8cd4bff5db14dda776ad4d2cba0962e80bbd6f 0 397 1725 2024-07-18T20:01:41Z Vrtnis 21 Created page with "= Robotic Control via Embodied Chain-of-Thought Reasoning = The codebase is built on top of OpenVLA. We refer to it for the detailed documentation of the code and dependencie..." wikitext text/x-wiki = Robotic Control via Embodied Chain-of-Thought Reasoning = The codebase is built on top of OpenVLA. We refer to it for the detailed documentation of the code and dependencies. == Quickstart == We provide a Colab notebook containing code for loading up our ECoT policy and using it to generate reasoning and actions in response to an observation. Loading the model for inference is easy: <code> from transformers import AutoModelForVision2Seq, AutoProcessor device = "cuda" path_to_hf = "Embodied-CoT/ecot-openvla-7b-bridge" processor = AutoProcessor.from_pretrained(path_to_hf, trust_remote_code=True) vla = AutoModelForVision2Seq.from_pretrained(path_to_hf, torch_dtype=torch.bfloat16).to(device) observation = &lt;ROBOT IMAGE OBSERVATION HERE&gt; instruction = &lt;YOUR INSTRUCTION HERE&gt; prompt = "A chat between a curious user and an artificial intelligence assistant. " + \ "The assistant gives helpful, detailed, and polite answers to the user's questions. " + \ f"USER: What action should the robot take to {instruction.lower()}? ASSISTANT: TASK:" inputs = processor(prompt, image).to(device, dtype=torch.bfloat16) action, generated_ids = vla.predict_action(**inputs, unnorm_key="bridge_orig", max_new_tokens=1024) generated_text = processor.batch_decode(generated_ids)[0] </code> The standard model in torch.bfloat16 requires 16 GB of GPU memory, but using bitsandbytes and 4-bit quantization lowers memory usage to around 5 GB. See the Colab for more details. == Training and Evaluation == To train the models from scratch, use the following command: <code> torchrun --standalone --nnodes 1 --nproc-per-node 8 vla-scripts/train.py \ --vla.type "prism-dinosiglip-224px+mx-bridge" \ --data_root_dir &lt;path to training data root&gt; \ --run_root_dir &lt;path to checkpoint saving directory&gt; \ --wandb_project &lt;wandb project name&gt; \ --wandb_entity &lt;wandb user name&gt; </code> To evaluate the model on the WidowX robot: <code> python3 experiments/bridge/eval_model_in_bridge_env.py --model.type prism-dinosiglip-224px+7b --pretrained_checkpoint &lt;path to checkpoint&gt; --host_ip &lt;robot interface IP&gt; --port &lt;robot interface port&gt; </code> == Pretrained models == We release two ECoT models trained as part of our work, and the dataset of reasonings, available on our HuggingFace page: * '''ecot-openvla-7b-bridge''': The main model that we used for most of our experiments. It was trained on the Bridge dataset annotated with the reasoning for 80k steps. * '''ecot-openvla-7b-oxe''': A policy that was initially trained on the Open-X-Embodiment dataset actions, fine-tuned on the mixture of OXE action-only data and our reasonings for Bridge for another 20k steps. * '''embodied_features_bridge''': A dataset of the embodied features and reasonings collected for Bridge demonstrations. === Explicit Notes on Model Licensing & Commercial Use === While all code in this repository is released under an MIT License, our pretrained models may inherit restrictions from the underlying base models we use. Specifically, both the above models are derived from Llama-2, and as such are subject to the Llama Community License. == Installation == See the original OpenVLA repository for detailed installation instructions. == Repository Structure == High-level overview of repository/project file-tree: * '''prismatic''': Package source; provides core utilities for model loading, training, data preprocessing, etc. * '''experiments''': Code for evaluating the policies on a WidowX robot. * '''vla-scripts/''': Core scripts for training, fine-tuning, and deploying VLAs. * '''LICENSE''': All code is made available under the MIT License; happy hacking! * '''Makefile''': Top-level Makefile (by default, supports linting - checking & auto-fix); extend as needed. * '''pyproject.toml''': Full project configuration details (including dependencies), as well as tool configurations. * '''README.md''': You are here! == Citation == If you find our code or models useful in your work, please cite our paper: <code> @article{Zawalski24-ecot, title={Robotic Control via Embodied Chain-of-Thought Reasoning}, author={Michał Zawalski and William Chen and Karl Pertsch and Oier Mees and Chelsea Finn and Sergey Levine}, journal={arXiv preprint arXiv:2407.08693}, year={2024} } </code> 6ff636270dbe3f66e88bfeb453ae2515e2342f9c 1726 1725 2024-07-18T20:05:49Z Vrtnis 21 wikitext text/x-wiki = Robotic Control via Embodied Chain-of-Thought Reasoning = Embodied Chain-of-Thought Reasoning (ECoT) is a novel approach for training robotic policies. This approach trains a vision-language-action model to generate reasoning steps in response to instructions and images before choosing a robot action, enabling better performance, interpretability, and generalization. The codebase is built on top of OpenVLA. Refer to it for detailed documentation of the code and dependencies. == Quickstart == A Colab notebook is provided containing code for loading the ECoT policy and using it to generate reasoning and actions in response to an observation. Loading the model for inference is easy: <code> from transformers import AutoModelForVision2Seq, AutoProcessor device = "cuda" path_to_hf = "Embodied-CoT/ecot-openvla-7b-bridge" processor = AutoProcessor.from_pretrained(path_to_hf, trust_remote_code=True) vla = AutoModelForVision2Seq.from_pretrained(path_to_hf, torch_dtype=torch.bfloat16).to(device) observation = &lt;ROBOT IMAGE OBSERVATION HERE&gt; instruction = &lt;YOUR INSTRUCTION HERE&gt; prompt = "A chat between a curious user and an artificial intelligence assistant. " + \ "The assistant gives helpful, detailed, and polite answers to the user's questions. " + \ f"USER: What action should the robot take to {instruction.lower()}? ASSISTANT: TASK:" inputs = processor(prompt, image).to(device, dtype=torch.bfloat16) action, generated_ids = vla.predict_action(**inputs, unnorm_key="bridge_orig", max_new_tokens=1024) generated_text = processor.batch_decode(generated_ids)[0] </code> The standard model in torch.bfloat16 requires 16 GB of GPU memory, but using bitsandbytes and 4-bit quantization lowers memory usage to around 5 GB. See the Colab for more details. == Training and Evaluation == To train the models from scratch, use the following command: <code> torchrun --standalone --nnodes 1 --nproc-per-node 8 vla-scripts/train.py \ --vla.type "prism-dinosiglip-224px+mx-bridge" \ --data_root_dir &lt;path to training data root&gt; \ --run_root_dir &lt;path to checkpoint saving directory&gt; \ --wandb_project &lt;wandb project name&gt; \ --wandb_entity &lt;wandb user name&gt; </code> To evaluate the model on the WidowX robot: <code> python3 experiments/bridge/eval_model_in_bridge_env.py --model.type prism-dinosiglip-224px+7b --pretrained_checkpoint &lt;path to checkpoint&gt; --host_ip &lt;robot interface IP&gt; --port &lt;robot interface port&gt; </code> == Pretrained models == Two ECoT models and a dataset of reasonings are available on the HuggingFace page: * '''ecot-openvla-7b-bridge''': The main model used for most of the experiments. It was trained on the Bridge dataset annotated with the reasoning for 80k steps. * '''ecot-openvla-7b-oxe''': A policy initially trained on the Open-X-Embodiment dataset actions, fine-tuned on the mixture of OXE action-only data and reasonings for Bridge for another 20k steps. * '''embodied_features_bridge''': A dataset of the embodied features and reasonings collected for Bridge demonstrations. === Explicit Notes on Model Licensing & Commercial Use === While all code in this repository is released under an MIT License, the pretrained models may inherit restrictions from the underlying base models. Specifically, both models are derived from Llama-2, and are subject to the Llama Community License. == Installation == Refer to the original OpenVLA repository for detailed installation instructions. == Repository Structure == High-level overview of repository/project file-tree: * '''prismatic''': Package source; provides core utilities for model loading, training, data preprocessing, etc. * '''experiments''': Code for evaluating the policies on a WidowX robot. * '''vla-scripts/''': Core scripts for training, fine-tuning, and deploying VLAs. * '''LICENSE''': All code is made available under the MIT License. * '''Makefile''': Top-level Makefile (by default, supports linting - checking & auto-fix); extend as needed. * '''pyproject.toml''': Full project configuration details (including dependencies), as well as tool configurations. * '''README.md''': You are here! == Citation == If the code or models are useful in your work, please cite the paper: <code> @article{Zawalski24-ecot, title={Robotic Control via Embodied Chain-of-Thought Reasoning}, author={Michał Zawalski and William Chen and Karl Pertsch and Oier Mees and Chelsea Finn and Sergey Levine}, journal={arXiv preprint arXiv:2407.08693}, year={2024} } </code> 5e6cd6c45c7c03e97f8d2fa94d421265aa0dfd99 0 275 1727 1186 2024-07-18T23:01:25Z Vrtnis 21 wikitext text/x-wiki Humanoid-Gym is an advanced reinforcement learning (RL) framework built on Nvidia Isaac Gym, designed for training locomotion skills in humanoid robots. Notably, it emphasizes zero-shot transfer, enabling skills learned in simulation to be directly applied to real-world environments without additional adjustments. [https://github.com/roboterax/humanoid-gym GitHub] == Demo == This codebase is verified by RobotEra's XBot-S (1.2 meter tall humanoid robot) and XBot-L (1.65 meter tall humanoid robot) in a real-world environment with zero-shot sim-to-real transfer. == Features == === 1. Humanoid Robot Training === This repository offers comprehensive guidance and scripts for the training of humanoid robots. Humanoid-Gym features specialized rewards for humanoid robots, simplifying the difficulty of sim-to-real transfer. In this repository, RobotEra's XBot-L is used as a primary example, but it can also be used for other robots with minimal adjustments. Resources cover setup, configuration, and execution, aiming to fully prepare the robot for real-world locomotion by providing in-depth training and optimization. * Comprehensive Training Guidelines: Thorough walkthroughs for each stage of the training process. * Step-by-Step Configuration Instructions: Clear and succinct guidance ensuring an efficient setup process. * Execution Scripts for Easy Deployment: Pre-prepared scripts to streamline the training workflow. === 2. Sim2Sim Support === Humanoid-Gym includes a sim2sim pipeline, allowing the transfer of trained policies to highly accurate and carefully designed simulated environments. Once the robot is acquired, the RL-trained policies can be confidently deployed in real-world settings. Simulator settings, particularly with Mujoco, are finely tuned to closely mimic real-world scenarios. This careful calibration ensures that performances in both simulated and real-world environments are closely aligned, making simulations more trustworthy and enhancing their applicability to real-world scenarios. === 3. Denoising World Model Learning (Coming Soon!) === Denoising World Model Learning (DWL) presents an advanced sim-to-real framework integrating state estimation and system identification. This dual-method approach ensures the robot's learning and adaptation are both practical and effective in real-world contexts. * Enhanced Sim-to-real Adaptability: Techniques to optimize the robot's transition from simulated to real environments. * Improved State Estimation Capabilities: Advanced tools for precise and reliable state analysis. === Dexterous Hand Manipulation (Coming Soon!) === == Installation == Generate a new Python virtual environment with Python 3.8: <code> conda create -n myenv python=3.8 </code> For best performance, it is recommended to use NVIDIA driver version 525: <code> sudo apt install nvidia-driver-525 </code> The minimal driver version supported is 515. If unable to install version 525, ensure that the system has at least version 515 to maintain basic functionality. Install PyTorch 1.13 with Cuda-11.7: <code> conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.7 -c pytorch -c nvidia </code> Install numpy-1.23: <code> conda install numpy=1.23 </code> Install Isaac Gym: Download and install Isaac Gym Preview 4 from [https://developer.nvidia.com/isaac-gym]. <code> cd isaacgym/python && pip install -e . </code> Run an example: <code> cd examples && python 1080_balls_of_solitude.py </code> Consult isaacgym/docs/index.html for troubleshooting. Install humanoid-gym: Clone this repository: <code> cd humanoid-gym && pip install -e . </code> == Usage Guide == === Examples === # Launching PPO Policy Training for 'v1' Across 4096 Environments <code> python scripts/train.py --task=humanoid_ppo --run_name v1 --headless --num_envs 4096 </code> # Evaluating the Trained PPO Policy 'v1' <code> python scripts/play.py --task=humanoid_ppo --run_name v1 </code> # Implementing Simulation-to-Simulation Model Transformation <code> python scripts/sim2sim.py --load_model /path/to/logs/XBot_ppo/exported/policies/policy_1.pt </code> # Run our trained policy <code> python scripts/sim2sim.py --load_model /path/to/logs/XBot_ppo/exported/policies/policy_example.pt </code> === 1. Default Tasks === * humanoid_ppo * Purpose: Baseline, PPO policy, Multi-frame low-level control * Observation Space: Variable dimensions, where is the number of frames * Privileged Information: Dimensions * humanoid_dwl (coming soon) === 2. PPO Policy === * Training Command: <code> python humanoid/scripts/train.py --task=humanoid_ppo --load_run log_file_path --name run_name </code> * Running a Trained Policy: <code> python humanoid/scripts/play.py --task=humanoid_ppo --load_run log_file_path --name run_name </code> By default, the latest model of the last run from the experiment folder is loaded. Other run iterations/models can be selected by adjusting load_run and checkpoint in the training config. === 3. Sim-to-sim === Before initiating the sim-to-sim process, ensure that play.py is run to export a JIT policy. Mujoco-based Sim2Sim Deployment: <code> python scripts/sim2sim.py --load_model /path/to/export/model.pt </code> === 4. Parameters === * CPU and GPU Usage: To run simulations on the CPU, set both --sim_device=cpu and --rl_device=cpu. For GPU operations, specify --sim_device=cuda:{0,1,2...} and --rl_device={0,1,2...} accordingly. Note that CUDA_VISIBLE_DEVICES is not applicable, and it's essential to match the --sim_device and --rl_device settings. * Headless Operation: Include --headless for operations without rendering. * Rendering Control: Press 'v' to toggle rendering during training. * Policy Location: Trained policies are saved in humanoid/logs/<experiment_name>/<date_time>_<run_name>/model_<iteration>.pt. === 5. Command-Line Arguments === For RL training, refer to humanoid/utils/helpers.py#L161. For the sim-to-sim process, refer to humanoid/scripts/sim2sim.py#L169. == Code Structure == Every environment hinges on an env file (legged_robot.py) and a configuration file (legged_robot_config.py). The latter houses two classes: LeggedRobotCfg (encompassing all environmental parameters) and LeggedRobotCfgPPO (denoting all training parameters). Both env and config classes use inheritance. Non-zero reward scales specified in cfg contribute a function of the corresponding name to the sum-total reward. Tasks must be registered with task_registry.register(name, EnvClass, EnvConfig, TrainConfig). Registration may occur within envs/__init__.py, or outside of this repository. === Add a new environment === The base environment legged_robot constructs a rough terrain locomotion task. The corresponding configuration does not specify a robot asset (URDF/MJCF) and no reward scales. If adding a new environment, create a new folder in the envs/ directory with a configuration file named <your_env>_config.py. The new configuration should inherit from existing environment configurations. === If proposing a new robot === * Insert the corresponding assets in the resources/ folder. * In the cfg file, set the path to the asset, define body names, default_joint_positions, and PD gains. Specify the desired train_cfg and the environment's name (python class). * In the train_cfg, set the experiment_name and run_name. * If needed, create your environment in <your_env>.py. Inherit from existing environments, override desired functions and/or add your reward functions. * Register your environment in humanoid/envs/__init__.py. * Modify or tune other parameters in your cfg or cfg_train as per requirements. To remove the reward, set its scale to zero. Avoid modifying the parameters of other environments! If a new robot/environment needs to perform sim2sim, modifications to humanoid/scripts/sim2sim.py may be required: * Check the joint mapping of the robot between MJCF and URDF. * Change the initial joint position of the robot according to the trained policy. === Troubleshooting === # ImportError: libpython3.8.so.1.0: cannot open shared object file: No such file or directory <code> export LD_LIBRARY_PATH="~/miniconda3/envs/your_env/lib:$LD_LIBRARY_PATH" </code> or <code> sudo apt install libpython3.8 </code> # AttributeError: module 'distutils' has no attribute 'version' <code> conda install pytorch torchvision torchaudio cudatoolkit=11.3 -c pytorch </code> # ImportError: /home/roboterax/anaconda3/bin/../lib/libstdc++.so.6: version `GLIBCXX_3.4.20` not found (required by /home/roboterax/carbgym/python/isaacgym/_bindings/linux64/gym_36.so) <code> mkdir ${YOUR_CONDA_ENV}/lib/_unused mv ${YOUR_CONDA_ENV}/lib/libstdc++* ${YOUR_CONDA_ENV}/lib/_unused </code> == Citation == Please cite the following if you use this code or parts of it: <code> @article{gu2024humanoid, title={Humanoid-Gym: Reinforcement Learning for Humanoid Robot with Zero-Shot Sim2Real Transfer}, author={Gu, Xinyang and Wang, Yen-Jen and Chen, Jianyu}, journal={arXiv preprint arXiv:2404.05695}, year={2024} } </code> == Acknowledgment == The implementation of Humanoid-Gym relies on resources from legged_gym and rsl_rl projects, created by the Robotic Systems Lab. The LeggedRobot implementation from their research is specifically utilized to enhance this codebase. e8b8c9faf0382ce4c4a66a5976f7abbf6098e1d0 0 398 1728 2024-07-20T21:50:50Z Vrtnis 21 Created page with "== Suggestions to Build MuJoCo WASM with Release 3.1.6 == 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com..." wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. e9212d658914d8a6d6ba3f289c556c3350ec1c49 1729 1728 2024-07-20T21:52:34Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 44278d2a4c04b6fca2eb34f9e7f28926dcb3af08 1730 1729 2024-07-21T00:22:47Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. d285a6d719a2065bb03898c9690ec249c083a11e 1731 1730 2024-07-21T00:32:45Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 48f845555806992ac0fce3fe1af85762267e1492 1732 1731 2024-07-21T00:36:14Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt: Build Options Modified The options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). Emscripten Setup Added New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Target Library Links Updated Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. cd252c5ede3f147aa94a694e255024daa9b84b64 1733 1732 2024-07-21T00:36:53Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Target Library Links Updated Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 055b6a12ee90709c9c28d4c96b78e804343c3cba 1734 1733 2024-07-21T00:37:21Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 7d5424b828be3f8791d4885d98244d526799988b 1735 1734 2024-07-21T00:39:25Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. Also, we'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 88bfd2bb9f05e519a470e9ee9ca178398b27cbb2 1736 1735 2024-07-21T00:39:46Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 900afaf152c5fb7cd51bc318f0025143bd8e90e5 1737 1736 2024-07-21T00:44:09Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> 8. '''Install the Build:''' <syntaxhighlight lang="sh"> cmake --install . sudo cmake --install . </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. a0244e7460286d4295fd24b30585718dec15f63a 6 399 1738 2024-07-21T01:16:12Z Vrtnis 21 Terminal output of wasm build wikitext text/x-wiki == Summary == Terminal output of wasm build 7030c3f83e987c7211e1dc67562e01f6b1fa6e82 0 398 1739 1737 2024-07-21T01:18:43Z Vrtnis 21 /*Add terminal output image*/ wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 361a859a694e9c897ddf922f7c6999c72e286ae8 1741 1739 2024-07-21T01:21:06Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 You probably have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 0551da456f4d57cfdeaaac83e06b4aad3703e4b9 1742 1741 2024-07-21T01:22:08Z Vrtnis 21 wikitext text/x-wiki == Suggestions to Build MuJoCo WASM with Release 3.1.6 == Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. b5322b45282b1f395c3703d4ac5a5b1e14e148eb 1743 1742 2024-07-21T01:23:05Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 4ed767b7fe954c2617e6b8275b54706fe811367b 1745 1743 2024-07-21T01:29:43Z Vrtnis 21 /*Add bin folder screenshot*/ wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/main...vrtnis:mujoco:release-3.1.6 As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. c1473f438dc492997ec1ef2832a820ea8dbcad5b 1746 1745 2024-07-21T02:09:09Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 04527b74dc62eee60f06658cc0ba0b83ab21d53b 1748 1746 2024-07-22T21:05:16Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show 5c4b86b b35d837 85f7539 6f173ea 4142aa5 7e2fb2f b0e4341 a4b173e ad49419 21e4e6c c18e6f5 836bbd9 534e43e 0b75f5f > changes.txt As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. f04c7bafd9bf4bc147b8692bca884d8d71341d68 1749 1748 2024-07-22T21:06:12Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show 5c4b86b b35d837 85f7539 6f173ea 4142aa5 7e2fb2f b0e4341 a4b173e ad49419 21e4e6c c18e6f5 836bbd9 534e43e 0b75f5f > changes.txt As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 39fae87acf8b07878a094751cec403db020899da 1750 1749 2024-07-22T21:26:54Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 394fa39d59c2fdbbba6a0fc400b2c6f5cbf4dbe1 1751 1750 2024-07-22T21:32:34Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. * For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. * * In src/user/user_objects.cc comment out the line that includes lodepng.h. Replace the bodies of the mjCHField::LoadPNG and mjCTexture::LoadPNG functions with a single return statement. * Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 9d8af48b64c7d5b97fc029c3e1cdec0ceb406929 1752 1751 2024-07-22T21:33:56Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. * For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. * * In src/user/user_objects.cc comment out the line that includes lodepng.h. Replace the bodies of the mjCHField::LoadPNG and mjCTexture::LoadPNG functions with a single return statement. * * In src/engine/engine_util_errmem.c update the preprocessor condition in the mju_writeLog function by replacing __STDC_VERSION_TIME_H__ with __EMSCRIPTEN__ in the #if directive. The line should now include __EMSCRIPTEN__ in the condition. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 55099cde94f01b807e8d29bb5e33f9b1cd8ccba5 1753 1752 2024-07-22T21:35:22Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. * For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. * * In src/user/user_objects.cc comment out the line that includes lodepng.h. Replace the bodies of the mjCHField::LoadPNG and mjCTexture::LoadPNG functions with a single return statement. * * In src/engine/engine_util_errmem.c update the preprocessor condition in the mju_writeLog function by replacing __STDC_VERSION_TIME_H__ with __EMSCRIPTEN__ in the #if directive. The line should now include __EMSCRIPTEN__ in the condition. * * In src/engine/engine_crossplatform.h add a conditional block specifically for Emscripten within the preprocessor directive in the engine_crossplatform.h file. Inside this block, include the sort_r.h header and define the mjQUICKSORT and quicksortfunc macros appropriately. This will separate the handling for Apple, Emscripten, and other platforms. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 776773235a1eb70a91503afc1e91cefbb44ae116 1754 1753 2024-07-22T21:36:19Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt ==== Specific ideas for changes ==== As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. * For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. * * In src/user/user_objects.cc comment out the line that includes lodepng.h. Replace the bodies of the mjCHField::LoadPNG and mjCTexture::LoadPNG functions with a single return statement. * * In src/engine/engine_util_errmem.c update the preprocessor condition in the mju_writeLog function by replacing __STDC_VERSION_TIME_H__ with __EMSCRIPTEN__ in the #if directive. The line should now include __EMSCRIPTEN__ in the condition. * * In src/engine/engine_crossplatform.h add a conditional block specifically for Emscripten within the preprocessor directive in the engine_crossplatform.h file. Inside this block, include the sort_r.h header and define the mjQUICKSORT and quicksortfunc macros appropriately. This will separate the handling for Apple, Emscripten, and other platforms. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 35deb54a2dddb4c4fe1a0ac44e3d52eedec1efe2 1755 1754 2024-07-22T21:37:43Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt ==== Specific ideas for changes ==== As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. * For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. * * In src/user/user_objects.cc comment out the line that includes lodepng.h. Replace the bodies of the mjCHField::LoadPNG and mjCTexture::LoadPNG functions with a single return statement. * * In src/engine/engine_util_errmem.c update the preprocessor condition in the mju_writeLog function by replacing __STDC_VERSION_TIME_H__ with __EMSCRIPTEN__ in the #if directive. The line should now include __EMSCRIPTEN__ in the condition. * * In src/engine/engine_crossplatform.h add a conditional block specifically for Emscripten within the preprocessor directive in the engine_crossplatform.h file. Inside this block, include the sort_r.h header and define the mjQUICKSORT and quicksortfunc macros appropriately. This will separate the handling for Apple, Emscripten, and other platforms. * * In the cmake/MujocoOptions.cmake file, remove the -Wno-int-in-bool-context compiler warning flag from the list of warnings. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. b2df887061079878f8bd4bcb429d72b61b672953 1756 1755 2024-07-22T21:39:08Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt ==== Specific ideas for changes ==== As of this writing, you'd need have to disable several features of 3.1.6 to allow for a successful WASM build. * For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. * * In src/user/user_objects.cc comment out the line that includes lodepng.h. Replace the bodies of the mjCHField::LoadPNG and mjCTexture::LoadPNG functions with a single return statement. * * In src/engine/engine_util_errmem.c update the preprocessor condition in the mju_writeLog function by replacing __STDC_VERSION_TIME_H__ with __EMSCRIPTEN__ in the #if directive. The line should now include __EMSCRIPTEN__ in the condition. * * In src/engine/engine_crossplatform.h add a conditional block specifically for Emscripten within the preprocessor directive in the engine_crossplatform.h file. Inside this block, include the sort_r.h header and define the mjQUICKSORT and quicksortfunc macros appropriately. This will separate the handling for Apple, Emscripten, and other platforms. * * In the cmake/MujocoOptions.cmake file, remove the -Wno-int-in-bool-context compiler warning flag from the list of warnings. * * In the CMakeLists.txt file, make the following changes: Change the default values of the options to disable the building of examples, simulate library, tests, and Python utility libraries by setting them to OFF: Set MUJOCO_BUILD_EXAMPLES to OFF. Set MUJOCO_BUILD_SIMULATE to OFF. Set MUJOCO_BUILD_TESTS to OFF. Set MUJOCO_TEST_PYTHON_UTIL to OFF. Remove the lodepng library from the target_link_libraries list for the mujoco target. Also, for example in mjxmacro.h you have to involve adding explicit casting to size_t for the calculations of key_mpos and key_mquat array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 6d3e1c65f1322f6abcd4cb6e68cb975fbac6b3b2 1759 1756 2024-07-22T21:43:58Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt ==== Specific ideas for changes ==== As of this writing, you'd need to disable several features of 3.1.6 to allow for a successful WASM build. For example in <pre>CMakeLists.txt</pre>, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. In `src/user/user_objects.cc`, comment out the line that includes `lodepng.h`. Replace the bodies of the `mjCHField::LoadPNG` and `mjCTexture::LoadPNG` functions with a single return statement. In `src/engine/engine_util_errmem.c`, update the preprocessor condition in the `mju_writeLog` function by replacing `__STDC_VERSION_TIME_H__` with `__EMSCRIPTEN__` in the `#if` directive. The line should now include `__EMSCRIPTEN__` in the condition. In `src/engine/engine_crossplatform.h`, add a conditional block specifically for Emscripten within the preprocessor directive. Inside this block, include the `sort_r.h` header and define the `mjQUICKSORT` and `quicksortfunc` macros appropriately. This will separate the handling for Apple, Emscripten, and other platforms. In the `cmake/MujocoOptions.cmake` file, remove the `-Wno-int-in-bool-context` compiler warning flag from the list of warnings. In the `CMakeLists.txt` file, make the following changes: * Change the default values of the options to disable the building of examples, simulate library, tests, and Python utility libraries by setting them to OFF: * Set `MUJOCO_BUILD_EXAMPLES` to OFF. * Set `MUJOCO_BUILD_SIMULATE` to OFF. * Set `MUJOCO_BUILD_TESTS` to OFF. * Set `MUJOCO_TEST_PYTHON_UTIL` to OFF. * Remove the `lodepng` library from the `target_link_libraries` list for the `mujoco` target. Also, for example, in `mjxmacro.h`, add explicit casting to `size_t` for the calculations of `key_mpos` and `key_mquat` array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 09e555c636883bc86d5a4407b89c693fe05b7baf 1760 1759 2024-07-22T21:44:13Z Vrtnis 21 /* Specific ideas for changes */ wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt ==== Specific ideas for changes ==== As of this writing, you'd need to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. In `src/user/user_objects.cc`, comment out the line that includes `lodepng.h`. Replace the bodies of the `mjCHField::LoadPNG` and `mjCTexture::LoadPNG` functions with a single return statement. In `src/engine/engine_util_errmem.c`, update the preprocessor condition in the `mju_writeLog` function by replacing `__STDC_VERSION_TIME_H__` with `__EMSCRIPTEN__` in the `#if` directive. The line should now include `__EMSCRIPTEN__` in the condition. In `src/engine/engine_crossplatform.h`, add a conditional block specifically for Emscripten within the preprocessor directive. Inside this block, include the `sort_r.h` header and define the `mjQUICKSORT` and `quicksortfunc` macros appropriately. This will separate the handling for Apple, Emscripten, and other platforms. In the `cmake/MujocoOptions.cmake` file, remove the `-Wno-int-in-bool-context` compiler warning flag from the list of warnings. In the `CMakeLists.txt` file, make the following changes: * Change the default values of the options to disable the building of examples, simulate library, tests, and Python utility libraries by setting them to OFF: * Set `MUJOCO_BUILD_EXAMPLES` to OFF. * Set `MUJOCO_BUILD_SIMULATE` to OFF. * Set `MUJOCO_BUILD_TESTS` to OFF. * Set `MUJOCO_TEST_PYTHON_UTIL` to OFF. * Remove the `lodepng` library from the `target_link_libraries` list for the `mujoco` target. Also, for example, in `mjxmacro.h`, add explicit casting to `size_t` for the calculations of `key_mpos` and `key_mquat` array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. 388c82fdc08cd136fe99f4c91b6aab897fc1a003 1765 1760 2024-07-22T22:29:29Z Vrtnis 21 wikitext text/x-wiki ==== Suggestions to Build MuJoCo WASM with Release 3.1.6 ==== Note: These instructions relate to a WASM build MuJoCo from the DeepMind MuJoCo source on Github. If you want to build using an existing WASM port of MuJoCo 2.3.1 check out [[MuJoCo_WASM]] 1. '''Clone the MuJoCo Repository:''' <syntaxhighlight lang="sh"> git clone --branch 3.1.6 https://github.com/deepmind/mujoco.git cd mujoco </syntaxhighlight> 2. '''Clone the Emscripten SDK Repository:''' <syntaxhighlight lang="sh"> git clone https://github.com/emscripten-core/emsdk.git cd emsdk </syntaxhighlight> 3. '''Install and Activate Emscripten:''' <syntaxhighlight lang="sh"> ./emsdk install latest ./emsdk activate latest source ./emsdk_env.sh cd .. </syntaxhighlight> Proceed with the next steps once you've made suggested changes to your codebase to allow for a WASM build. You can see the changes here as a reference https://github.com/google-deepmind/mujoco/compare/3.1.6...vrtnis:mujoco:release-3.1.6 You could also run git show to see specific changes that have been made to the original mujoco 3.1.6. files e.g. ,5c4b86b > changes.txt ==== Specific ideas for changes ==== As of this writing, you'd need to disable several features of 3.1.6 to allow for a successful WASM build. For example in CMakeLists.txt, the options to build examples, simulate library, and tests for MuJoCo have been turned off by default (changed from ON to OFF). New configurations and target properties for building MuJoCo with Emscripten have been added. This includes defining source files, checking their existence, and setting specific properties and options for building WebAssembly (.wasm) and HTML output. Some target link options have been adjusted, including removing the lodepng library from the target link list. * `src/user/user_objects.cc`, comment out the line that includes `lodepng.h`. Replace the bodies of the `mjCHField::LoadPNG` and `mjCTexture::LoadPNG` functions with a single return statement. * * `src/engine/engine_util_errmem.c`, update the preprocessor condition in the `mju_writeLog` function by replacing `__STDC_VERSION_TIME_H__` with `__EMSCRIPTEN__` in the `#if` directive. The line should now include `__EMSCRIPTEN__` in the condition. * * `src/engine/engine_crossplatform.h`, add a conditional block specifically for Emscripten within the preprocessor directive. Inside this block, include the `sort_r.h` header and define the `mjQUICKSORT` and `quicksortfunc` macros appropriately. This will separate the handling for Apple, Emscripten, and other platforms. * * `cmake/MujocoOptions.cmake` file, remove the `-Wno-int-in-bool-context` compiler warning flag from the list of warnings. In the `CMakeLists.txt` file, make the following changes: * Change the default values of the options to disable the building of examples, simulate library, tests, and Python utility libraries by setting them to OFF: * Set `MUJOCO_BUILD_EXAMPLES` to OFF. * Set `MUJOCO_BUILD_SIMULATE` to OFF. * Set `MUJOCO_BUILD_TESTS` to OFF. * Set `MUJOCO_TEST_PYTHON_UTIL` to OFF. * Remove the `lodepng` library from the `target_link_libraries` list for the `mujoco` target. Also, for example, in `mjxmacro.h`, add explicit casting to `size_t` for the calculations of `key_mpos` and `key_mquat` array sizes, ensuring correct memory allocation and preventing potential integer overflow issues. We'd suggest taking a look at https://github.com/stillonearth/MuJoCo-WASM/issues/1 (older 2.3.1 build but still relevant) 4. '''Prepare the Build Environment:''' <syntaxhighlight lang="sh"> mkdir build cd build </syntaxhighlight> 5. '''Run Emscripten CMake Commands:''' <syntaxhighlight lang="sh"> emcmake cmake .. emmake make </syntaxhighlight> [[File:Screenshot18-10-40.png|WASM Build Terminal Output]] 6. '''Deploy and Run Locally:''' <syntaxhighlight lang="sh"> emrun --no_browser --port 8080 . </syntaxhighlight> [[File:Screenshot18-10-07.png|800px|WASM Build Running In Browser]] [[File:Screenshot18-24-43.png|800px|WASM Build bin folder]] 7. '''Optional Cleanup and Repeat Steps if Necessary:''' <syntaxhighlight lang="sh"> rm -rf * emcmake cmake .. emmake make </syntaxhighlight> === Notes === * Ensure that the Emscripten environment is correctly activated before starting the build process. * Regularly clean the build directory to maintain a clean build environment. abf1626a092569dca4a43dcb5bc49a0dce9e5993 6 400 1740 2024-07-21T01:19:23Z Vrtnis 21 Emscripten output wikitext text/x-wiki == Summary == Emscripten output 63f1db9e5527b61b33cae69772dc3ca23d3fc5c7 6 401 1744 2024-07-21T01:25:40Z Vrtnis 21 bin folder after successful wasm build wikitext text/x-wiki == Summary == bin folder after successful wasm build 675f4a9e7693866fbc7268c5e6356bd69fb3ab8e 0 299 1747 1722 2024-07-22T15:29:52Z Ben 2 wikitext text/x-wiki {| class="wikitable" |- ! Date ! Presenter ! Topic |- | 2024.07.21 | Wesley | Consistency Models |- | 2024.07.20 | Nathan | Flow Matching |- | 2024.07.17 | Dennis | Diffusion Models |- | 2024.07.16 | Isaac | Convolutional Neural Networks |- | 2024.07.15 | Allen | RNNs, LSTMs, GRUs |- | 2024.06.28 | Kenji | Principles of Power Electronics |- | 2024.06.27 | Paweł | OpenVLA |- | 2024.06.26 | Ryan | Introduction to KiCAD |- | 2024.06.24 | Dennis | Live Streaming Protocols |- | 2024.06.21 | Nathan | Quantization |- | 2024.06.20 | Timothy | Diffusion |- | 2024.06.19 | Allen | Neural Network Loss Functions |- | 2024.06.18 | Ben | Neural network inference on the edge |- | 2024.06.17 | Matt | CAD software deep dive |- | 2024.06.14 | Allen | [https://github.com/karpathy/llm.c llm.c] |- | 2024.06.13 | Kenji | Engineering principles of BLDCs |- | 2024.06.12 | Ryan | 3-phase motors |- | 2024.06.11 | Vedant | CAN Protocol |- | 2024.06.10 | Isaac | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech Papers Round 2] |- | 2024.06.07 | Ben | Quantization |- | 2024.06.06 | Tom | Linux Raw |- | 2024.06.05 | Hugo | Gaussian Splats |- | 2024.06.04 | Dennis | [https://humanoids.wiki/w/Dennis%27_Speech_Project Speech representation learning papers] |- | 2024.06.03 | Paweł | What I (want to) believe in |- | 2024.05.30 | Isaac | VLMs |- | 2024.05.29 | Allen | PPO |} [[Category: K-Scale]] 68f2c7d55fe9e8c254e8d16432a5c061e78d5e05 6 402 1757 2024-07-22T21:42:29Z Ben 2 wikitext text/x-wiki KawaiiStompySticker2 69972620ddc1080c0f3e76ed4ce90ae798fa45e1 0 2 1758 1664 2024-07-22T21:42:39Z Ben 2 /* Artwork */ wikitext text/x-wiki [[File:Stompy.jpg|right|300px|thumb|Stompy standing up]] {{infobox robot | name = Stompy | organization = [[K-Scale Labs]] | cost = USD 10,000 }} Stompy is an open-source humanoid robot developed by [[K-Scale Labs]]. Here are some relevant links: * [[Stompy To-Do List]] * [[Stompy Build Guide]] * [[Gripper History]] = Hardware = This page is dedicated to detailing the hardware selections for humanoid robots, including various components such as actuators, cameras, compute units, PCBs and modules, batteries, displays, microphones, speakers, as well as wiring and connectors. == Actuators == Actuators are the components that allow the robot to move and interact with its environment. They convert energy into mechanical motion. Common types used in humanoid robots include: * Servo motors * Stepper motors * Linear actuators == Cameras == Cameras are essential for visual processing, allowing the robot to perceive its surroundings. Important considerations include: * Resolution and frame rate * Field of view * Depth sensing capabilities (3D cameras) == Compute == The compute section handles the processing requirements of the robot. This includes: * Microprocessors and microcontrollers * Single-board computers like Raspberry Pi or Nvidia Jetson * Dedicated AI accelerators for machine learning tasks == PCB and Modules == Printed Circuit Boards (PCBs) and the modules on them are the backbone of the robot's electronic system. * Main control board * Power management modules * Sensor interfaces * Communication modules (Wi-Fi, Bluetooth) == Batteries == Batteries provide the necessary power to all robotic systems and are crucial for mobile autonomy. Selection factors include: * Battery type (Li-Ion, NiMH, Lead-Acid) * Capacity (measured in mAh or Ah) * Voltage and energy density * Safety features and durability == Displays == Displays are used in robots for displaying information such as system status, data, and interactive elements. Key features include: * Size variations ranging from small to large panels * Touchscreen capabilities * High resolution displays == Microphones == Microphones enable the robot to receive and process audio inputs, crucial for voice commands and auditory data. Factors to consider are: * Sensitivity and noise cancellation * Directionality (omnidirectional vs. unidirectional) * Integration with voice recognition software == Speakers == Speakers allow the robot to communicate audibly with its environment, essential for interaction and alerts. Considerations include: * Power output and sound quality * Size and mounting options * Compatibility with audio processing hardware == Wiring and Connectors == Proper wiring and connectors ensure reliable communication and power supply throughout the robot's components. * Types of wires (gauge, shielding) * Connectors (pin types, waterproofing) * Cable management solutions === Conventions === The images below show our pin convention for the CAN bus when using various connectors. <gallery> Kscale db9 can bus convention.jpg Kscale phoenix can bus convention.jpg </gallery> = Simulation = For the latest simulation artifacts, see [https://kscale.dev/ the website]. = Artwork = Here's some art of Stompy! <gallery> Stompy 1.png Stompy 2.png Stompy 3.png Stompy 4.png KawaiiStompySticker2.png </gallery> [[Category:Robots]] [[Category:Open Source]] [[Category:K-Scale]] 5bd60361744d0590c8a07ffa6bd67befe837d581 0 403 1761 2024-07-22T21:52:10Z Vrtnis 21 Created page with "= Notes from user calls == - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, ment..." wikitext text/x-wiki = Notes from user calls == - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json 8b23643e23c746a1519e63b1fdf6678488081d9e 1762 1761 2024-07-22T21:52:29Z Vrtnis 21 /* Notes from user calls = */ wikitext text/x-wiki ==== Notes from user calls ==== - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json c8d28c7ec53a33c7cc33325cd18d7efe6cc0563d 1763 1762 2024-07-22T21:52:43Z Vrtnis 21 wikitext text/x-wiki ==== Notes from user calls ==== - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json 1fe0f7d320c7520c7c114f349b19a2c37c31c3e5 1764 1763 2024-07-22T21:55:13Z Vrtnis 21 wikitext text/x-wiki ==== Notes from user calls ==== Work in progress... - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json dd32c242095403f3195056a7a388fcf0cad7df90 1767 1764 2024-07-22T22:52:34Z Vrtnis 21 wikitext text/x-wiki ==== Notes from user calls ==== Work in progress... - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system '''Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json''' [[File:Screenshot15-50-46.png|900px]] a8d22e1d994c3f123cb6574c8321dba4bc58e2f6 1768 1767 2024-07-22T22:52:54Z Vrtnis 21 wikitext text/x-wiki ==== Notes from user calls ==== ''Work in progress... '' - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system '''Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json''' [[File:Screenshot15-50-46.png|900px]] 42e1c8a590ab97806b660739fb62b5fed0626c70 1769 1768 2024-07-22T22:53:04Z Vrtnis 21 wikitext text/x-wiki ==== Notes from user calls ==== ''Work in progress...'' - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system '''Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json''' [[File:Screenshot15-50-46.png|900px]] 598c2f0e534eb5c5bde1bfde426a69d103f97d07 1770 1769 2024-07-22T22:53:17Z Vrtnis 21 wikitext text/x-wiki ==== Notes from user calls ==== - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system '''Source transcript at: https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json''' [[File:Screenshot15-50-46.png|900px]] 76fb32912827b6430784261856175f48238f1994 1771 1770 2024-07-22T22:53:45Z Vrtnis 21 wikitext text/x-wiki ==== Notes from user calls ==== - how robotics simulators have changed processes and improved workflow. - using MuJoCo for advanced simulations and engine design performance, mentioning issues with communicating these simulations to engineering and maintenance teams. - open source simulation tools, such as creating a comprehensive simulation platform, exploring engine dynamics, and the ease of customization due to its open-source nature. - example use case of MuJoCo providing self-serve exploration for maintenance teams, refining designs, developing predictive maintenance strategies, and reducing development costs. - self-serve simulations work and their user-friendly design for maintenance personnel. - customized KUKA robot and a visual inspection system using a cobot for automated documentation of engine parts. - robotic polishing system '''Source transcript at:''' === https://vrtnis.github.io/transcript-viewer/?jsonUrl=https://gist.githubusercontent.com/vrtnis/3a025a21cbe367368b01fe8b97d8d9d9/raw/1f2962a16f7b6439d36f8067c8f1657a72408682/pratt_1-1948273645.json === [[File:Screenshot15-50-46.png|900px]] 516d6b4478a2db25923f0692151a52846531f5fc 6 404 1766 2024-07-22T22:51:18Z Vrtnis 21 Call transcript screenshot wikitext text/x-wiki == Summary == Call transcript screenshot 0897791029faf53701958eb8f3362adc8ca37938 0 1 1772 1705 2024-07-24T20:00:24Z Ben 2 added a couple actuators wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === List of Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === List of Actuators === {| class="wikitable" |- ! Actuator ! Notes |- | [[OBot]] | Open-source actuator |- | [[SPIN Servo]] | Open-source actuator |- | [[VESCular6]] | A project based on [[VESC]] |- | [[ODrive]] | A precision motor controller |- | [[Solo Motor Controller]] | A motor controller alternative to the [[ODrive]]. |- | [[J60]] | Actuators built for the [[DEEP Robotics]] quadrupeds. |- | [[Robstride]] | |- | [[MyActuator]] | |- | [[K-Scale Motor Controller]] | An open-source motor controller |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] 63e8146dc2a86888bb4779f188ff0497fe9d2d43 1773 1772 2024-07-24T20:05:03Z Ben 2 change sections around, split actuators and motor controllers wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[J60]] |- | [[Robstride]] |- | [[MyActuator]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] 0a90d29b98382679269406f36929a7ff2af5eb76 1776 1773 2024-07-24T20:05:28Z Ben 2 /* Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] 7ad6d2dc9ac1edb7d25f048a7743e755186889ea 1783 1776 2024-07-24T20:11:14Z Ben 2 /* Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] fdd120e83941ecb213f3af630ccdab1dbb721c85 0 197 1774 816 2024-07-24T20:05:19Z Ben 2 Ben moved page [[J60]] to [[DEEP Robotics J60]] wikitext text/x-wiki The [https://www.deeprobotics.cn/en/index/j60.html J60] by [[DEEP Robotics]] is a family of high-performance actuators designed for use in quadruped and humanoid robots. == Overview == The J60 actuators are high-performance actuators have been specifically designed for use in humanoid robots and in [[DEEP Robotics]] quadrupeds. These actuators are reliable and durable, featuring a high torque-to-weight ratio. The joints include a gear reduction, a frameless torque motor, a servo driver, and an absolute value encoder into one compact unit. === Actuators === {{infobox actuator | name = J60-6 | manufacturer = DEEP Robotics | link = https://www.deeprobotics.cn/en/index/j60.html | peak_torque = 19.94 Nm | peak_speed = 24.18 rad/s | dimensions = 76.5mm diameter, 63mm length | weight = 480g | absolute_encoder_resolution = 14 bit | operating_voltage_range = 12-36V | standard_operating_voltage = 24V | interface = CAN bus / RS485 | control_frequency = 1kHz }} {{infobox actuator | name = J60-10 | manufacturer = DEEP Robotics | link = https://www.deeprobotics.cn/en/index/j60.html | peak_torque = 30.50 Nm | peak_speed = 15.49 rad/s | dimensions = 76.5mm diameter, 72.5mm length | weight = 540g | absolute_encoder_resolution = 14 bit | operating_voltage_range = 12-36V | standard_operating_voltage = 24V | interface = CAN bus / RS485 | control_frequency = 1kHz }} [[Category: Actuators]] b5cdee73845d295957f9422b4ae2f30130be176c 1777 1774 2024-07-24T20:05:41Z Ben 2 /* Actuators */ wikitext text/x-wiki The [https://www.deeprobotics.cn/en/index/j60.html J60] by [[DEEP Robotics]] is a family of high-performance actuators designed for use in quadruped and humanoid robots. == Overview == The J60 actuators are high-performance actuators have been specifically designed for use in humanoid robots and in [[DEEP Robotics]] quadrupeds. These actuators are reliable and durable, featuring a high torque-to-weight ratio. The joints include a gear reduction, a frameless torque motor, a servo driver, and an absolute value encoder into one compact unit. === Actuators === {{infobox actuator | name = J60-6 | manufacturer = DEEP Robotics | link = https://www.deeprobotics.cn/en/index/j60.html | peak_torque = 19.94 Nm | peak_speed = 24.18 rad/s | dimensions = 76.5mm diameter, 63mm length | weight = 480g | absolute_encoder_resolution = 14 bit | operating_voltage_range = 12-36V | standard_operating_voltage = 24V | interface = CAN bus / RS485 | control_frequency = 1kHz }} {{infobox actuator | name = J60-10 | manufacturer = DEEP Robotics | link = https://www.deeprobotics.cn/en/index/j60.html | peak_torque = 30.50 Nm | peak_speed = 15.49 rad/s | dimensions = 76.5mm diameter, 72.5mm length | weight = 540g | absolute_encoder_resolution = 14 bit | operating_voltage_range = 12-36V | standard_operating_voltage = 24V | interface = CAN bus / RS485 | control_frequency = 1kHz }} [[Category: Actuators]] b2bcdb453bf0f1c0a488c65cfede379b5ca03e6d 0 405 1775 2024-07-24T20:05:19Z Ben 2 Ben moved page [[J60]] to [[DEEP Robotics J60]] wikitext text/x-wiki #REDIRECT [[DEEP Robotics J60]] c07b4fe6c14364c1abf7935d2d63b9bb5c72064e 0 406 1778 2024-07-24T20:06:54Z Ben 2 Created page with " {{infobox actuator | name = RobStride 01 | manufacturer = RobStride | link = https://robstride.com/products/robStride01 | peak_torque = 17 Nm | peak_speed = | dimensions = |..." wikitext text/x-wiki {{infobox actuator | name = RobStride 01 | manufacturer = RobStride | link = https://robstride.com/products/robStride01 | peak_torque = 17 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | operating_voltage_range = | standard_operating_voltage = | interface = CAN | control_frequency = }} 600f2c0ef2c11943d47de206fc6ee56ac52d46cf 1779 1778 2024-07-24T20:08:18Z Ben 2 wikitext text/x-wiki [https://robstride.com/ Robstride] is an actuator manufacturing startup. The company was founded in 2024 by founders from Xiaomi that worked on the CyberDog and CyberGear projects. == Actutators == {{infobox actuator | name = RobStride 01 | manufacturer = RobStride | link = https://robstride.com/products/robStride01 | peak_torque = 17 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | operating_voltage_range = | standard_operating_voltage = | interface = CAN | control_frequency = }} 701a0bb15d21f90d5e0770ab376cf93a40e9d295 1780 1779 2024-07-24T20:08:46Z Ben 2 /* Actutators */ wikitext text/x-wiki [https://robstride.com/ Robstride] is an actuator manufacturing startup. The company was founded in 2024 by founders from Xiaomi that worked on the CyberDog and CyberGear projects. == Actutators == {{infobox actuator | name = RobStride 01 | manufacturer = RobStride | link = https://robstride.com/products/robStride01 | peak_torque = 17 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | operating_voltage_range = | standard_operating_voltage = | interface = CAN | control_frequency = }} {{infobox actuator | name = RobStride 04 | manufacturer = RobStride | link = https://robstride.com/products/robStride04 | peak_torque = 120 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | operating_voltage_range = | standard_operating_voltage = | interface = CAN | control_frequency = }} d95d3df3e14ee60e1682130520325b3a4e95af2f 1781 1780 2024-07-24T20:09:51Z Ben 2 /* Actutators */ wikitext text/x-wiki [https://robstride.com/ Robstride] is an actuator manufacturing startup. The company was founded in 2024 by founders from Xiaomi that worked on the CyberDog and CyberGear projects. == Actutators == {{infobox actuator | name = RobStride 01 | manufacturer = RobStride | purchase_link = https://robstride.com/products/robStride01 | peak_torque = 17 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | operating_voltage_range = | standard_operating_voltage = | interface = CAN | control_frequency = }} {{infobox actuator | name = RobStride 04 | manufacturer = RobStride | purchase_link = https://robstride.com/products/robStride04 | peak_torque = 120 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | operating_voltage_range = | standard_operating_voltage = | interface = CAN | control_frequency = }} a93cdab5b7555e84cd309aba28be6f326ad02420 1782 1781 2024-07-24T20:10:25Z Ben 2 /* Actutators */ wikitext text/x-wiki [https://robstride.com/ Robstride] is an actuator manufacturing startup. The company was founded in 2024 by founders from Xiaomi that worked on the CyberDog and CyberGear projects. == Actutators == {{infobox actuator | name = RobStride 01 | manufacturer = RobStride | purchase_link = https://robstride.com/products/robStride01 | peak_torque = 17 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = }} {{infobox actuator | name = RobStride 04 | manufacturer = RobStride | purchase_link = https://robstride.com/products/robStride04 | peak_torque = 120 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = }} 4dbf89e031740926480092778c5a3966e5f659a0 6 407 1784 2024-07-24T20:15:45Z Ben 2 wikitext text/x-wiki EC-A10020-P1-12-6 39da5e4a9f286ecc28e7b933c023c101dee15695 0 408 1785 2024-07-24T20:15:55Z Ben 2 Created page with "[http://encos.cn/ Encos] is a Chinese actuator manufacturer {{infobox actuator | name = EC-A10020-P1-12/6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.a..." wikitext text/x-wiki [http://encos.cn/ Encos] is a Chinese actuator manufacturer {{infobox actuator | name = EC-A10020-P1-12/6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=183 | nominal_torque = 50 Nm | peak_torque = 150 Nm | peak_speed = | dimensions = | weight = 1.35 Kg | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A10020-P1-12-6.png|thumb]] 3f822a7914d47e1f1f582e5358a611692d146357 6 409 1786 2024-07-24T20:17:24Z Ben 2 wikitext text/x-wiki EC-A8112-P1-18 ba3d328a1a5125ccba90523ea28032e852b51dd7 6 410 1787 2024-07-24T20:18:22Z Ben 2 wikitext text/x-wiki EC-A4310-P2-36 0e15c25d87939644dec63563af8b7991c0da388c 6 411 1788 2024-07-24T20:19:45Z Ben 2 wikitext text/x-wiki EC-A8120-P1-6 cb53eb7e8cde55ea6ff823422aaf19598f44dcb6 0 408 1789 1785 2024-07-24T20:19:57Z Ben 2 add more actuators wikitext text/x-wiki [http://encos.cn/ Encos] is a Chinese actuator manufacturer {{infobox actuator | name = EC-A10020-P1-12/6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=183 | nominal_torque = 50 Nm | peak_torque = 150 Nm | peak_speed = | dimensions = | weight = 1.35 Kg | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A10020-P1-12-6.png|thumb|none|EC-A10020-P1-12/6]] {{infobox actuator | name = EC-A8112-P1-18 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=203 | nominal_torque = 30 Nm | peak_torque = 90 Nm | peak_speed = | dimensions = | weight = 840 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A8112-P1-18.png|thumb|none|EC-A8112-P1-18]] {{infobox actuator | name = EC-A8112-P1-18 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=204 | nominal_torque = 30 Nm | peak_torque = 90 Nm | peak_speed = | dimensions = | weight = 840 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A4310-P2-36.png|thumb|none|EC-A4310-P2-36]] {{infobox actuator | name = EC-A8120-P1-6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=205 | nominal_torque = 15 Nm | peak_torque = 50 Nm | peak_speed = | dimensions = | weight = 890 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A8120-P1-6.png|thumb|none|EC-A8120-P1-6]] 43b3c5567dab4fa7ff608b70580df28d01e46253 1790 1789 2024-07-24T20:21:46Z Ben 2 correct information, move images around wikitext text/x-wiki [http://encos.cn/ Encos] is a Chinese actuator manufacturer [[File:EC-A10020-P1-12-6.png|thumb|right|EC-A10020-P1-12/6]] {{infobox actuator | name = EC-A10020-P1-12/6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=183 | nominal_torque = 50 Nm | peak_torque = 150 Nm | peak_speed = | dimensions = | weight = 1.35 Kg | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A8112-P1-18.png|thumb|right|EC-A8112-P1-18]] {{infobox actuator | name = EC-A8112-P1-18 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=203 | nominal_torque = 30 Nm | peak_torque = 90 Nm | peak_speed = | dimensions = | weight = 840 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A4310-P2-36.png|thumb|right|EC-A4310-P2-36]] {{infobox actuator | name = EC-A4310-P2-36 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=204 | nominal_torque = 12 Nm | peak_torque = 36 Nm | peak_speed = | dimensions = | weight = 377 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A8120-P1-6.png|thumb|right|EC-A8120-P1-6]] {{infobox actuator | name = EC-A8120-P1-6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=205 | nominal_torque = 15 Nm | peak_torque = 50 Nm | peak_speed = | dimensions = | weight = 890 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} f3c3f352522eb5bd835f8166f7515cb5a023fab7 1795 1790 2024-07-24T20:35:36Z Ben 2 wikitext text/x-wiki [http://encos.cn/ Encos] is a Chinese actuator manufacturer. [[File:EC-A10020-P1-12-6.png|thumb|right|EC-A10020-P1-12/6]] {{infobox actuator | name = EC-A10020-P1-12/6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=183 | nominal_torque = 50 Nm | peak_torque = 150 Nm | peak_speed = | dimensions = | weight = 1.35 Kg | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A8112-P1-18.png|thumb|right|EC-A8112-P1-18]] {{infobox actuator | name = EC-A8112-P1-18 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=203 | nominal_torque = 30 Nm | peak_torque = 90 Nm | peak_speed = | dimensions = | weight = 840 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A4310-P2-36.png|thumb|right|EC-A4310-P2-36]] {{infobox actuator | name = EC-A4310-P2-36 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=204 | nominal_torque = 12 Nm | peak_torque = 36 Nm | peak_speed = | dimensions = | weight = 377 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} [[File:EC-A8120-P1-6.png|thumb|right|EC-A8120-P1-6]] {{infobox actuator | name = EC-A8120-P1-6 | manufacturer = Encos | purchase_link = http://encos.cn/ProDetail.aspx?ProID=205 | nominal_torque = 15 Nm | peak_torque = 50 Nm | peak_speed = | dimensions = | weight = 890 g | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = 2000 Hz }} 1dd58848943c3e1c055e6ad4d230d927ada7a71f 0 1 1791 1783 2024-07-24T20:28:32Z Ben 2 /* Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] c72fe58381bfc5362eed2b82ce1e73d57e74676d 1808 1791 2024-08-13T18:43:44Z Ben 2 /* Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/rhCy6UdBRD Discord] fd610b0436105b3fbd13757cea15dfa0c07efd7f 1825 1808 2024-08-29T18:46:41Z Ben 2 /* Discord community */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] 14960d96ca5c8a9180d5a63d9851a55f7926e115 6 412 1792 2024-07-24T20:31:19Z Ben 2 wikitext text/x-wiki Steadywin-gim10015-9 0e5df7b138e2e14659e3aa0c57d565acf7167885 0 413 1793 2024-07-24T20:31:31Z Ben 2 Created page with "[http://steadywin.cn/en/ Steadywin] is a Chinese actuator manufacturer [[File:Steadywin-gim10015-9.jpg|thumb|right|GIM10015-9]] {{infobox actuator | name = GIM10015-9 | manu..." wikitext text/x-wiki [http://steadywin.cn/en/ Steadywin] is a Chinese actuator manufacturer [[File:Steadywin-gim10015-9.jpg|thumb|right|GIM10015-9]] {{infobox actuator | name = GIM10015-9 | manufacturer = Steadywin | purchase_link = http://steadywin.cn/en/pd.jsp?id=114#_jcp=3_3 | nominal_torque = 28 Nm | peak_torque = 70 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | voltage = 48V | interface = CAN / RS485 | control_frequency = }} 135b0567e8b756d85dbaec3b299c6f6f834542e8 1794 1793 2024-07-24T20:34:26Z Ben 2 wikitext text/x-wiki [http://steadywin.cn/en/ Steadywin] is a Chinese actuator manufacturer [[File:Steadywin-gim10015-9.jpg|thumb|right|GIM10015-9]] {{infobox actuator | name = GIM10015-9 | manufacturer = Steadywin | purchase_link = http://steadywin.cn/en/pd.jsp?id=114#_jcp=3_3 | nominal_torque = 28 Nm | peak_torque = 70 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | voltage = 48V | interface = CAN / RS485 | control_frequency = }} {{infobox actuator | name = GIM8115-36 | manufacturer = Steadywin | purchase_link = http://steadywin.cn/en/pd.jsp?id=19#_jcp=3_3 | nominal_torque = 54 Nm | peak_torque = 150 Nm | peak_speed = | dimensions = | weight = | absolute_encoder_resolution = | voltage = 48V | interface = CAN | control_frequency = }} d0bdd55561fc88847d6367cb3e96fedee96aea57 0 414 1796 2024-07-24T21:28:03Z Vrtnis 21 /*create page*/ wikitext text/x-wiki '''Suzhou Micro Actuator Technology Co., Ltd''' (MyActuator) specializes in high-performance brushless servo motors, ideal for a range of applications including robots and portable devices. Their products, noted for their compact yet powerful design, offer versatility and precision, making them a top choice for robotics enthusiasts and professionals alike. 0cd7b178c0fd7aa1e3ed6c4b56ed648ff5485d2d 1797 1796 2024-07-24T21:29:46Z Vrtnis 21 /*add RMD-X6-P8-8*/ wikitext text/x-wiki '''Suzhou Micro Actuator Technology Co., Ltd''' (MyActuator) specializes in high-performance brushless servo motors, ideal for a range of applications including robots and portable devices. Their products, noted for their compact yet powerful design, offer versatility and precision, making them a top choice for robotics enthusiasts and professionals alike. {{infobox actuator | name = RMD-X6-P8-8 | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | cost = USD 1 | purchase_link = https://www.myactuator.com/x6-8-details | nominal_torque = 8 Nm | peak_torque = 8 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | interface = CAN | gear_type = Planetary }} df0d9ffee825b9ed9b206febef4fdd7abad204c3 1798 1797 2024-07-24T21:33:00Z Vrtnis 21 wikitext text/x-wiki '''Suzhou Micro Actuator Technology Co., Ltd''' (MyActuator) specializes in high-performance brushless servo motors, ideal for a range of applications including robots and portable devices. Their products, noted for their compact yet powerful design, offer versatility and precision, making them a top choice for robotics enthusiasts and professionals alike. {{infobox actuator | name = RMD-X6-P8-8 | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x6-8-details | nominal_torque = 8 Nm | peak_torque = 8 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | interface = CAN | gear_type = Planetary }} 0a6cef1d1d8050714d2987b835ebe719b6e7f647 1799 1798 2024-07-24T22:08:27Z Vrtnis 21 /*RMD-X8-P6-20-C-N*/ wikitext text/x-wiki '''Suzhou Micro Actuator Technology Co., Ltd''' (MyActuator) specializes in high-performance brushless servo motors, ideal for a range of applications including robots and portable devices. Their products, noted for their compact yet powerful design, offer versatility and precision, making them a top choice for robotics enthusiasts and professionals alike. {{infobox actuator | name = RMD-X6-P8-8 | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x6-8-details | nominal_torque = 8 Nm | peak_torque = 8 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | interface = CAN | gear_type = Planetary }} {{infobox actuator | name = RMD-X8-P6-20-C-N | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x8-20-details | peak_torque = 20 Nm | gear_ratio = 1:6 | interface = CAN | gear_type = Planetary }} 1e72aa6e1accda6487fc728827c0483d401585b0 1800 1799 2024-07-24T22:14:30Z Vrtnis 21 wikitext text/x-wiki '''Suzhou Micro Actuator Technology Co., Ltd''' (MyActuator) specializes in high-performance brushless servo motors, ideal for a range of applications including robots and portable devices. Their products, noted for their compact yet powerful design, offer versatility and precision, making them a top choice for robotics enthusiasts and professionals alike. {{infobox actuator | name = RMD-X6-P8-8 | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x6-8-details | nominal_torque = 8 Nm | peak_torque = 8 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | interface = CAN | gear_type = Planetary }} {{infobox actuator | name = RMD-X8-P6-20-C-N | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x8-20-details | peak_torque = 20 Nm | gear_ratio = 1:6 | interface = CAN | gear_type = Planetary }} {{infobox actuator | name = RMD-X8-P36-60-C-N | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x8-60-details | peak_torque = 60 Nm | gear_ratio = 1:36 | interface = CAN | gear_type = Planetary }} ef1b8d5a5970c8829579e972e8ba4ec0df6ad372 1801 1800 2024-07-24T22:29:56Z Vrtnis 21 wikitext text/x-wiki '''Suzhou Micro Actuator Technology Co., Ltd''' (MyActuator) specializes in high-performance brushless servo motors, ideal for a range of applications including robots and portable devices. Their products, noted for their compact yet powerful design, offer versatility and precision, making them a top choice for robotics enthusiasts and professionals alike. {{infobox actuator | name = RMD-X6-P8-8 | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x6-8-details | nominal_torque = 8 Nm | peak_torque = 8 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | interface = CAN | gear_type = Planetary }} {{infobox actuator | name = RMD-X8-P6-20-C-N | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x8-20-details | peak_torque = 20 Nm | gear_ratio = 1:6 | interface = CAN | gear_type = Planetary }} {{infobox actuator | name = RMD-X8-P36-60-C-N | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x8-60-details | peak_torque = 60 Nm | gear_ratio = 1:36 | interface = CAN | gear_type = Planetary }} [[Category:Companies]] e6ab3243426e6888f8015184537ac0e1b3205ead 1803 1801 2024-07-24T22:38:06Z Vrtnis 21 /* Add HQ image*/ wikitext text/x-wiki '''Suzhou Micro Actuator Technology Co., Ltd''' (MyActuator) specializes in high-performance brushless servo motors, ideal for a range of applications including robots and portable devices. Their products, noted for their compact yet powerful design, offer versatility and precision, making them a top choice for robotics enthusiasts and professionals alike. [[File:Myactuator hq.jpg|thumb|MyActuator HQ Suzhou]] {{infobox actuator | name = RMD-X6-P8-8 | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x6-8-details | nominal_torque = 8 Nm | peak_torque = 8 Nm | weight = 1 kg | dimensions = 10cm radius | gear_ratio = 1:8 | voltage = 48V | interface = CAN | gear_type = Planetary }} {{infobox actuator | name = RMD-X8-P6-20-C-N | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x8-20-details | peak_torque = 20 Nm | gear_ratio = 1:6 | interface = CAN | gear_type = Planetary }} {{infobox actuator | name = RMD-X8-P36-60-C-N | manufacturer = Suzhou Micro Actuator Technology Co., Ltd. | purchase_link = https://www.myactuator.com/x8-60-details | peak_torque = 60 Nm | gear_ratio = 1:36 | interface = CAN | gear_type = Planetary }} [[Category:Companies]] 8d8515e2dff1fb4f610c9f5c3a0d84e04f509325 6 415 1802 2024-07-24T22:36:55Z Vrtnis 21 MyActuator HQ image wikitext text/x-wiki == Summary == MyActuator HQ image 0400c6bbfdc428c45afd5448335d5ceae3d28f8f 0 294 1804 1720 2024-07-26T16:02:44Z Ben 2 wikitext text/x-wiki [[File:Robot taking notes.png|thumb|Robot taking notes (from Midjourney)]] {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1815523839407006189 2024.07.19] |- | [https://x.com/kscalelabs/status/1811805432505336073 2024.07.12] |- | [https://x.com/kscalelabs/status/1809263616958374286 2024.07.05] |- | [https://x.com/kscalelabs/status/1804184936574030284 2024.06.21] |- | [https://x.com/kscalelabs/status/1801749382167204086 2024.06.14] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] 12e10bd1d266c0884a8633dac1af996d136fe898 1807 1804 2024-08-02T16:49:11Z Ben 2 wikitext text/x-wiki [[File:Robot taking notes.png|thumb|Robot taking notes (from Midjourney)]] {| class="wikitable" |- ! Link |- | [https://x.com/kscalelabs/status/1819413752539935164 2024.08.02] |- | [https://x.com/kscalelabs/status/1816987089386553543 2024.07.26] |- | [https://x.com/kscalelabs/status/1815523839407006189 2024.07.19] |- | [https://x.com/kscalelabs/status/1811805432505336073 2024.07.12] |- | [https://x.com/kscalelabs/status/1809263616958374286 2024.07.05] |- | [https://x.com/kscalelabs/status/1804184936574030284 2024.06.21] |- | [https://x.com/kscalelabs/status/1801749382167204086 2024.06.14] |- | [https://x.com/kscalelabs/status/1799197382208590132 2024.06.07] |- | [https://x.com/kscalelabs/status/1796617681455775944 2024.05.31] |- | [https://x.com/kscalelabs/status/1794109131214712914 2024.05.24] |- | [https://x.com/kscalelabs/status/1791507358780461496 2024.05.17] |- | [https://x.com/kscalelabs/status/1788968705378181145 2024.05.10] |} [[Category:K-Scale]] 02b05fc865118fb3b04ebcb8894034132adb1b41 0 416 1805 2024-08-01T10:29:34Z Ben 2 openscad notes wikitext text/x-wiki [https://openscad.org/ OpenSCAD] is a programmatic CAD tool. There are some nice extensions to make dealing with OpenSCAD less painful: - [https://github.com/BelfrySCAD/BOSL2 BOSL2], which stands for the Belfry OpenSCAD Library, is a collection of implementations of OpenSCAD components. - [https://pypi.org/project/solidpython2/ SolidPython2] is a Python frontend for OpenSCAD which converts to the OpenSCAD DSL. 69f56a42425e465996093786b342e58cce18e371 1806 1805 2024-08-01T10:30:01Z Ben 2 fix lists wikitext text/x-wiki [https://openscad.org/ OpenSCAD] is a programmatic CAD tool. There are some nice extensions to make dealing with OpenSCAD less painful: * [https://github.com/BelfrySCAD/BOSL2 BOSL2], which stands for the Belfry OpenSCAD Library, is a collection of implementations of OpenSCAD components. * [https://pypi.org/project/solidpython2/ SolidPython2] is a Python frontend for OpenSCAD which converts to the OpenSCAD DSL. 3d13c6df39bf774ca3a0eb4f7359b93bc8908601 0 417 1809 2024-08-13T18:49:03Z Ben 2 Created page with "[https://www.elementalmotors.com/ Elemental Motors] is building high-torque direct drive motors." wikitext text/x-wiki [https://www.elementalmotors.com/ Elemental Motors] is building high-torque direct drive motors. d5de29643dcea63a95b3eecb8e73eefba4f9f28b 1810 1809 2024-08-13T18:55:15Z Ben 2 add motors wikitext text/x-wiki [https://www.elementalmotors.com/ Elemental Motors] is building high-torque direct drive motors. {{infobox actuator | name = G4 57 Inner Rotator | manufacturer = Elemental Motors | nominal_torque = 3.4 Nm | peak_torque = 5.43 Nm | peak_speed = 930 RPM | dimensions = OD: 57mm, ID: 22mm, Length: 23mm | weight = 0.25 kg | voltage = 48V | interface = N/A | control_frequency = N/A }} {{infobox actuator | name = G3 114 12T Inner Rotator | manufacturer = Elemental Motors | nominal_torque = 25 Nm | peak_torque = 53 Nm | peak_speed = 627 RPM | dimensions = OD: 114mm, ID: 60mm, Length: 38.8mm | weight = 1.5 kg | voltage = 48V | interface = N/A | control_frequency = N/A }} {{infobox actuator | name = G4 95 Inner Rotator | manufacturer = Elemental Motors | nominal_torque = 12.6 Nm | peak_torque = 35.4 Nm | peak_speed = 253 RPM | dimensions = OD: 95mm, ID: 45mm, Length: 38.5mm | weight = 1.26 kg | voltage = 48V | interface = N/A | control_frequency = N/A }} {{infobox actuator | name = G4 115 42T Inner Rotator | manufacturer = Elemental Motors | nominal_torque = 35.7 Nm | peak_torque = 74.1 Nm | peak_speed = 155 RPM | dimensions = OD: 115mm, ID: 60mm, Length: 46mm | weight = 2.1 kg | voltage = 48V | interface = N/A | control_frequency = N/A }} {{infobox actuator | name = G4 170 Inner Rotator | manufacturer = Elemental Motors | nominal_torque = 108 Nm | peak_torque = 212 Nm | peak_speed = 50 RPM | dimensions = OD: 170mm, ID: 106mm, Length: 60mm | weight = 4 kg | voltage = 48V | interface = N/A | control_frequency = N/A }} 9b9e9eade14f10fb3b0d5facd6982bbf65c5db2b 0 10 1811 216 2024-08-17T06:34:03Z Joshc 75 updated overview and history wikitext text/x-wiki [https://www.1x.tech/ 1X] (formerly known as Halodi Robotics) is a humanoid robotics company based in Moss, Norway. The company was founded in 2014 with the mission to create an abundant supply of labor via safe, intelligent robots.<ref>https://www.1x.tech/about</ref> They have two robots: [[Eve]] and [[Neo]]. Eve was the first generation, a wheeled robot designed to train their AI models and collect real-world data. Neo is their second model, an improved, bipedal version of Eve which is still under development. Eve and Neo are designed for safe human interaction by reducing actuator inertia. The goal for Eve is to work in logistics and guarding, while Neo is intended to be used in the home. {{infobox company | name = 1X Technologies | country = United States | website_link = https://www.1x.tech/ | robots = [[Eve]], [[Neo]] }} The company is known for its high torque brushless direct current (BLDC) motor Revo1 that they developed in house. Those BLDC motors are paired with low gear ratio cable drives, and are currently the highest torque-to-weight direct drive motor in the world. == History == In 2020, 1X partnered with Everon by ADT Commercial, a commercial security solutions company, to deploy 150-250 humanoid robots in various buildings across the US for night guarding. This is a unique use-case for humanoids because other companies such as Figure and Tesla have been focusing on warehouse and factory applications. In March 2023, the company closed a funding round of $23.5 million, led by the OpenAI Startup Fund, along with other investors including Tiger Global, Sandwater, Alliance Ventures, and Skagerak Capital.<ref>https://robotsguide.com/robots/eve</ref> The partnership with OpenAI has helped significantly for developing the AI system used in their humanoids. In January 2024, 1X announced that they raised a $100 million Series B led by EQT Ventures, including Samsung NEXT and existing investors Tiger Global and Nistad Group.<ref>https://www.therobotreport.com/1x-technologies-raises-100m-series-b-advance-neo-humanoid-robot</ref> == References == [[Category:Companies]] 15508cb60101814803001c9ad192b1b341804508 1813 1811 2024-08-17T17:13:24Z Joshc 75 wikitext text/x-wiki [https://www.1x.tech/ 1X] (formerly known as Halodi Robotics) is a humanoid robotics company based in Moss, Norway. The company was founded in 2014 with the mission to create an abundant supply of labor via safe, intelligent robots.<ref>https://www.1x.tech/about</ref> They have two robots: [[Eve]] and [[Neo]]. Eve was the first generation, a wheeled robot designed to train their AI models and collect real-world data. Neo is their second model, an improved, bipedal version of Eve which is still under development. Eve and Neo are designed for safe human interaction by reducing actuator inertia. The goal for Eve is to work in logistics and guarding, while Neo is intended to be used in the home. {{infobox company | name = 1X Technologies | country = United States | website_link = https://www.1x.tech/ | robots = [[Eve]], [[Neo]] }} The company is known for its high torque brushless direct current (BLDC) motor Revo1 that they developed in house. Those BLDC motors are paired with low gear ratio cable drives, and are currently the highest torque-to-weight direct drive motor in the world. == History == In 2020, 1X partnered with Everon by ADT Commercial, a commercial security solutions company, to deploy 150-250 humanoid robots in various buildings across the US for night guarding. In March 2023, the company closed a funding round of $23.5 million, led by the OpenAI Startup Fund, along with other investors including Tiger Global, Sandwater, Alliance Ventures, and Skagerak Capital.<ref>https://robotsguide.com/robots/eve</ref> The partnership with OpenAI has helped significantly for developing the AI system used in their humanoids. In January 2024, 1X announced that they raised a $100 million Series B led by EQT Ventures, including Samsung NEXT and existing investors Tiger Global and Nistad Group.<ref>https://www.therobotreport.com/1x-technologies-raises-100m-series-b-advance-neo-humanoid-robot</ref> == References == [[Category:Companies]] 42eb70ad70b772d124ad5ba4997c544d80886496 0 54 1812 1640 2024-08-17T06:45:03Z Joshc 75 wikitext text/x-wiki EVE is a versatile and agile humanoid robot developed by [[1X]]. It is equipped with cameras and sensors to perceive and interact with its surroundings. EVE’s mobility, dexterity, and balance allow it to navigate complex environments and manipulate objects effectively. <youtube>https://www.youtube.com/watch?v=20GHG-R9eFI</youtube> {{infobox robot | name = EVE | organization = [[1X]] | height = 186 cm | weight = 86 kg | speed = 14.4 km/hr | carry_capacity = 15 kg | runtime = 6 hrs | video_link = https://www.youtube.com/watch?v=20GHG-R9eFI }} == Specifications == Eve is designed with a single leg and a set of wheels, and grippers for the hands. It has a total of 25 degrees of freedom (DOF): 1 in the neck, 7 in each arm, 6 in the leg, and 1 in each hand and the wheels. It is made mostly out of plastic, aluminum, and fabric. It stands at 183cm tall (the same height as the founder and CEO Bernt Øivind Børnich), weighs 83lbs, and moves at a top speed of 14.4km/h. The computer runs on an Intel i7 CPU for real-time processing and Nvidia Xavier for AI computations. It uses a custom Linux-based operating system, Java, C++, Python, and many open-source packages. Eve has 4 hours of operating time on a 1.05-kWh lithium-ion battery pack.<ref>https://robotsguide.com/robots/eve</ref> == References == [[Category:Robots]] 03dbaffeb703b443be13fccce65b90711db41def 0 418 1814 2024-08-20T22:20:49Z Ben 2 Created page with "These are notes for the MinPPO project [https://github.com/kscalelabs/minppo here]. == Testing == * Hidden layer size of 256 shows progress (loss is based on state.q[2]) * s..." wikitext text/x-wiki These are notes for the MinPPO project [https://github.com/kscalelabs/minppo here]. == Testing == * Hidden layer size of 256 shows progress (loss is based on state.q[2]) * setting std to zero makes rewards nans why. I wonder if there NEEDS to be randomization in the enviornment * ctrl cost is whats giving nans? interesting? * it is unrelated to randomization of enviornmnet. i think gradient related * first thing to become nans seems to be actor loss and scores. after that, everything becomes nans * fixed entropy epsilon. hope this works now. b4b3cea66b44aa913c98eed1580a14fe113dbd89 1815 1814 2024-08-20T22:21:30Z Ben 2 wikitext text/x-wiki These are notes for the MinPPO project [https://github.com/kscalelabs/minppo here]. == Testing == * Hidden layer size of 256 shows progress (loss is based on state.q[2]) * setting std to zero makes rewards nans why. I wonder if there NEEDS to be randomization in the enviornment * ctrl cost is whats giving nans? interesting? * it is unrelated to randomization of enviornmnet. i think gradient related * first thing to become nans seems to be actor loss and scores. after that, everything becomes nans * fixed entropy epsilon. hope this works now. == MNIST Training Example == <syntaxhighlight lang="python"> """Trains a simple MNIST model using Equinox.""" import logging import equinox as eqx import jax import jax.numpy as jnp import optax from jax import random from tensorflow.keras.datasets import mnist logging.basicConfig( level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s", datefmt="%Y-%m-%d %H:%M:%S" ) logger = logging.getLogger(__name__) def load_mnist() -> tuple[jnp.ndarray, jnp.ndarray, jnp.ndarray, jnp.ndarray]: """Load and preprocess the MNIST dataset.""" (x_train, y_train), (x_test, y_test) = mnist.load_data() x_train = jnp.float32(x_train) / 255.0 x_test = jnp.float32(x_test) / 255.0 x_train = x_train.reshape(-1, 28 * 28) x_test = x_test.reshape(-1, 28 * 28) y_train = jax.nn.one_hot(y_train, 10) y_test = jax.nn.one_hot(y_test, 10) return x_train, y_train, x_test, y_test class DenseModel(eqx.Module): """Define a simple dense neural network model.""" layers: list def __init__(self, key: jnp.ndarray) -> None: keys = random.split(key, 3) self.layers = [ eqx.nn.Linear(28 * 28, 128, key=keys[0]), eqx.nn.Linear(128, 64, key=keys[1]), eqx.nn.Linear(64, 10, key=keys[2]), ] def __call__(self, x: jnp.ndarray) -> jnp.ndarray: for layer in self.layers[:-1]: x = jax.nn.relu(layer(x)) return self.layers[-1](x) @eqx.filter_value_and_grad def loss_fn(model: DenseModel, x_b: jnp.ndarray, y_b: jnp.ndarray) -> jnp.ndarray: """Define the loss function (cross-entropy loss). Vecotrized across batch with vmap.""" pred_b = jax.vmap(model)(x_b) return -jnp.mean(jnp.sum(y_b * jax.nn.log_softmax(pred_b), axis=-1)) @eqx.filter_jit def make_step( model: DenseModel, optimizer: optax.GradientTransformation, opt_state: optax.OptState, x_b: jnp.ndarray, y_b: jnp.ndarray, ) -> tuple[jnp.ndarray, DenseModel, optax.OptState]: """Perform a single optimization step.""" loss, grads = loss_fn(model, x_b, y_b) updates, opt_state = optimizer.update(grads, opt_state) model = eqx.apply_updates(model, updates) return loss, model, opt_state def train( model: DenseModel, optimizer: optax.GradientTransformation, x_train: jnp.ndarray, y_train: jnp.ndarray, batch_size: int, num_epochs: int, ) -> DenseModel: """Train the model using the given data.""" opt_state = optimizer.init(eqx.filter(model, eqx.is_array)) for epoch in range(num_epochs): for i in range(0, len(x_train), batch_size): x_batch = x_train[i : i + batch_size] y_batch = y_train[i : i + batch_size] loss, model, opt_state = make_step(model, optimizer, opt_state, x_batch, y_batch) logger.info(f"Epoch {epoch+1}/{num_epochs}, Loss: {loss:.4f}") return model @eqx.filter_jit def accuracy(model: DenseModel, x_b: jnp.ndarray, y_b: jnp.ndarray) -> jnp.ndarray: """Takes in batch oftest images/label pairing with model and returns accuracy.""" pred = jax.vmap(model)(x_b) return jnp.mean(jnp.argmax(pred, axis=-1) == jnp.argmax(y_b, axis=-1)) def main() -> None: # Load data x_train, y_train, x_test, y_test = load_mnist() # Initialize model and optimizer key = random.PRNGKey(0) model = DenseModel(key) optimizer = optax.adam(learning_rate=1e-3) # Train the model batch_size = 32 num_epochs = 10 trained_model = train(model, optimizer, x_train, y_train, batch_size, num_epochs) test_accuracy = accuracy(trained_model, x_test, y_test) logger.info(f"Test accuracy: {test_accuracy:.4f}") if __name__ == "__main__": main() </syntaxhighlight> d52e06e64c08148fb8e6bc4c0c63f68855f3a062 0 419 1816 2024-08-23T02:24:36Z Goblinrum 25 Created page with "For controlling multiple CAN buses, our solution is to use the Infineon XMC4800, a MCU that integrates a M4 core and has both an EtherCAT node and 6x CAN nodes. This allows us..." wikitext text/x-wiki For controlling multiple CAN buses, our solution is to use the Infineon XMC4800, a MCU that integrates a M4 core and has both an EtherCAT node and 6x CAN nodes. This allows us to control the CAN buses through either FS USB or (in the future) EtherCAT. ==== Testing the Evaluation Board ==== === Programming the MCU === ==== CAN example code ==== === Integrated XMC4800 CAN board === TBD [[Category:Guides]] [[Category:Electronics]] [[Category:Hardware]] 577a14779bb1e63c176440312861ded7a0301c28 1818 1816 2024-08-23T02:37:02Z Goblinrum 25 wikitext text/x-wiki For controlling multiple CAN buses, our solution is to use the Infineon XMC4800, a MCU that integrates a M4 core and has both an EtherCAT node and 6x CAN nodes. This allows us to control the CAN buses through either FS USB or (in the future) EtherCAT. * [https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/ Product Page] * [https://www.infineon.com/dgdl/Infineon-XMC4700-XMC4800-ReferenceManual-v01_02-EN.pdf?fileId=5546d46250cc1fdf01513f8e052d07fc Eval Board Reference Manual] * [https://www.infineon.com/dgdl/Infineon-XMC4700-XMC4800-DataSheet-v01_02-EN.pdf?fileId=5546d462518ffd850151908ea8db00b3 Data Sheet] * [https://www.infineon.com/dgdl/Infineon-xmc4800-UM-v01_00-EN.pdf?fileId=5546d4624d6fc3d5014ddd85f6d97832 MCU Reference Manual] Most technical implementation details are in the Reference Manual ==== Testing the Evaluation Board ==== [[File:XMC4800 Breadboard.png|thumb]] The testing setup is connected as the image describes. The eval board has one built-in CAN transceiver (while the rest of them are using external SN65HVD230 modules). The one built-in transceiver is connected to the other four CAN buses as a debug RX path === Programming the MCU === Follow the instructions [https://github.com/Infineon/mtb-example-xmc-can-loopback here] for an example on how to program the XMC4800 chip. The main takeaway is # You need the J-SEGGER driver for your computer # Connect via the DEBUG micro-B port, not the other USB port # Use ModusToolBox to get started with the code examples ==== CAN example code ==== @Tom + @Ved(?) TODO === Integrated XMC4800 CAN board === TBD [[Category:Guides]] [[Category:Electronics]] [[Category:Hardware]] 8fdb46f03e876bcd31135af35d3988e7d991b4f9 1819 1818 2024-08-23T03:07:20Z Goblinrum 25 /* Testing the Evaluation Board */ wikitext text/x-wiki For controlling multiple CAN buses, our solution is to use the Infineon XMC4800, a MCU that integrates a M4 core and has both an EtherCAT node and 6x CAN nodes. This allows us to control the CAN buses through either FS USB or (in the future) EtherCAT. * [https://www.infineon.com/cms/en/product/microcontroller/32-bit-industrial-microcontroller-based-on-arm-cortex-m/32-bit-xmc4000-industrial-microcontroller-arm-cortex-m4/xmc4800/ Product Page] * [https://www.infineon.com/dgdl/Infineon-XMC4700-XMC4800-ReferenceManual-v01_02-EN.pdf?fileId=5546d46250cc1fdf01513f8e052d07fc Eval Board Reference Manual] * [https://www.infineon.com/dgdl/Infineon-XMC4700-XMC4800-DataSheet-v01_02-EN.pdf?fileId=5546d462518ffd850151908ea8db00b3 Data Sheet] * [https://www.infineon.com/dgdl/Infineon-xmc4800-UM-v01_00-EN.pdf?fileId=5546d4624d6fc3d5014ddd85f6d97832 MCU Reference Manual] Most technical implementation details are in the Reference Manual ==== Testing the Evaluation Board ==== [[File:XMC4800 Breadboard.png|thumb]] The testing setup is connected as the image describes. The eval board has one built-in CAN transceiver (while the rest of them are using external SN65HVD230 modules). The one built-in transceiver is connected to the other four CAN buses as a debug RX path. This should be disconnected for testing otherwise since the built in bus is only meant to do basic loopback-style testing === Programming the MCU === Follow the instructions [https://github.com/Infineon/mtb-example-xmc-can-loopback here] for an example on how to program the XMC4800 chip. The main takeaway is # You need the J-SEGGER driver for your computer # Connect via the DEBUG micro-B port, not the other USB port # Use ModusToolBox to get started with the code examples ==== CAN example code ==== @Tom + @Ved(?) TODO === Integrated XMC4800 CAN board === TBD [[Category:Guides]] [[Category:Electronics]] [[Category:Hardware]] 8de28cb3c91934dd4fdfda548acd794953d98857 6 420 1817 2024-08-23T02:29:56Z Goblinrum 25 wikitext text/x-wiki Breadboarded wiring for testing the Eval Board 22d38301f338178e885878506039851c72b1e304 0 315 1820 1541 2024-08-28T06:29:58Z Vedant 24 wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: https://developer.nvidia.com/sdk-manager # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = # Repeat Steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel,</code> and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release of the downloadable sources, enter <code>jetson_36.3</code>. This will install the f13369375fc8dccd2653e361090b54e0e2bf63e4 1821 1820 2024-08-28T06:44:43Z Vedant 24 /* Flashing Custom Firmware (For Jetson 36.3) */ wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: https://developer.nvidia.com/sdk-manager # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. == Downloading the Toolchain == == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. 8b8cccf6a95d78bdf3bb3a140bc451ad81010a5f 1822 1821 2024-08-28T06:49:30Z Vedant 24 /* Downloading the Toolchain */ wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: https://developer.nvidia.com/sdk-manager # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. 0dec29ec46946ccbe62ce4e6a3501dd2786580b9 1823 1822 2024-08-28T18:45:54Z Vedant 24 wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: https://developer.nvidia.com/sdk-manager # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the 3e90ac814bb42582c504f56e8e7a7b985939a726 1826 1823 2024-08-29T22:42:52Z Vedant 24 wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: https://developer.nvidia.com/sdk-manager # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the <code>kernel</code> eventually navigate to the <code>kernel-jammy-src</code> folder and run <code>make menuconfig ARCH=arm64</code>. This will bring up a UI with configurable drivers and peripherals. Select desired configurations and save. # The configurations can be found within a <code>.config</code> file located within the same directory. Copy the contents and locate the <code>defconfig</code> file in <code>./arch/arm64/configs/</code>, overwriting it with the copied contents. # ccedfbdf13919f720e9dc4ecb69e936fb102c363 1827 1826 2024-08-29T23:04:43Z Vedant 24 wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: https://developer.nvidia.com/sdk-manager # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the <code>kernel</code> eventually navigate to the <code>kernel-jammy-src</code> folder and run <code>make menuconfig ARCH=arm64</code>. This will bring up a UI with configurable drivers and peripherals. Select desired configurations and save. # The configurations can be found within a <code>.config</code> file located within the same directory. Copy the contents and locate the <code>defconfig</code> file in <code>./arch/arm64/configs/</code>, overwriting it with the copied contents. == Building Custom Kernel and Installing Modules == # Navigate back out to <code>sources</code>. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE=<toolchain-path>/bin/aarch64-buildroot-linux-gnu-</code>. If installation was done correctly as per the pre-requisites section, the command <code>export CROSS_COMPILE=$HOME/l4t-gcc/aarch64--glibc--stable-2022.08-1/bin/aarch64-buildroot-linux-gnu-</code> should work. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE_AARCH64_PATH=</code>, and <code>export CROSS_COMPILE_AARCH64=/bin/aarch64-buildroot-linux-gnu-</code>. (Potentially deprecated) # Inside the sources, directory, make an output directory for built kernel files using <code>mkdir kernel_out</code>. # Build the modules using the command <code>./nvbuild.sh -o kernel_out</code>. # 2ddb0f02245307be74e80d8038f10deff16ce282 1828 1827 2024-08-30T19:42:22Z Vedant 24 wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: <code>https://developer.nvidia.com/sdk-manager</code> # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the <code>kernel</code> eventually navigate to the <code>kernel-jammy-src</code> folder and run <code>make menuconfig ARCH=arm64</code>. This will bring up a UI with configurable drivers and peripherals. Select desired configurations and save. # The configurations can be found within a <code>.config</code> file located within the same directory. Copy the contents and locate the <code>defconfig</code> file in <code>./arch/arm64/configs/</code>, overwriting it with the copied contents. == Building Custom Kernel and Installing Modules == # Navigate back out to <code>sources</code>. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE=<toolchain-path>/bin/aarch64-buildroot-linux-gnu-</code>. If installation was done correctly as per the pre-requisites section, the command <code>export CROSS_COMPILE=$HOME/l4t-gcc/aarch64--glibc--stable-2022.08-1/bin/aarch64-buildroot-linux-gnu-</code> should work. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE_AARCH64_PATH=</code>, and <code>export CROSS_COMPILE_AARCH64=/bin/aarch64-buildroot-linux-gnu-</code>. (Potentially deprecated) # Inside the sources, directory, make an output directory for built kernel files using <code>mkdir kernel_out</code>. # Build the modules using the command <code>./nvbuild.sh -o kernel_out</code>. This will compile the drivers and device trees for the new kernel. # Navigate out from the <code>sources</code> directory into the <code>Linux for Tegra</code>. # Use the <code>cp</code> to overwrite <code>./rootfs/usr/lib/modules/5.15.136-tegra/updates/nvgpu.ko</code> with <code>./source/kernel_out/nvgpu/drivers/gpu/nvgpu/nvgpu.ko</code>. # Repeat the previous step to replace <code>Linux_for_Tegra/kernel/dtb/</code> with # To specify the installation path for the compiled modules, use the command <code>export INSTALL_MOD_PATH=$HOME/nvidia/nvidia_sdk/JetPack_6.0_Linux_JETSON_AGX_ORIN_TARGETS/Linux_for_Tegra/rootfs/</code>. # Install modules using the command <code>./nvbuild -i</code>. == Flashing the Kernel == # Ensure that c7fe2c7630ebe3f9ee5e661d97fb8788aac82c50 1829 1828 2024-08-30T19:54:28Z Vedant 24 wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: <code>https://developer.nvidia.com/sdk-manager</code> # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the <code>kernel</code> eventually navigate to the <code>kernel-jammy-src</code> folder and run <code>make menuconfig ARCH=arm64</code>. This will bring up a UI with configurable drivers and peripherals. Select desired configurations and save. # The configurations can be found within a <code>.config</code> file located within the same directory. Copy the contents and locate the <code>defconfig</code> file in <code>./arch/arm64/configs/</code>, overwriting it with the copied contents. == Building Custom Kernel and Installing Modules == # Navigate back out to <code>sources</code>. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE=<toolchain-path>/bin/aarch64-buildroot-linux-gnu-</code>. If installation was done correctly as per the pre-requisites section, the command <code>export CROSS_COMPILE=$HOME/l4t-gcc/aarch64--glibc--stable-2022.08-1/bin/aarch64-buildroot-linux-gnu-</code> should work. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE_AARCH64_PATH=</code>, and <code>export CROSS_COMPILE_AARCH64=/bin/aarch64-buildroot-linux-gnu-</code>. (Potentially deprecated) # Inside the sources, directory, make an output directory for built kernel files using <code>mkdir kernel_out</code>. # Build the modules using the command <code>./nvbuild.sh -o kernel_out</code>. This will compile the drivers and device trees for the new kernel. # Navigate out from the <code>sources</code> directory into the <code>Linux for Tegra</code>. # Use the <code>cp</code> to overwrite <code>./rootfs/usr/lib/modules/5.15.136-tegra/updates/nvgpu.ko</code> with <code>./source/kernel_out/nvgpu/drivers/gpu/nvgpu/nvgpu.ko</code>. # Repeat the previous step to replace <code>Linux_for_Tegra/kernel/dtb/</code> with </code>source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/dts/nvidia</code>. Ensure that instead of overwriting the directory, only the files are copied over. # Overwrite the Image in <code>./kernel</code> with <code>./source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/Image </code>. # To specify the installation path for the compiled modules, use the command <code>export INSTALL_MOD_PATH=$HOME/nvidia/nvidia_sdk/JetPack_6.0_Linux_JETSON_AGX_ORIN_TARGETS/Linux_for_Tegra/rootfs/</code>. # Install modules using the command <code>./nvbuild -i</code>. The Jetson is now ready to be flashed. == Flashing the Kernel == Ensure that the target Jetson is connected to the host device and is in recovery mode. Navigate to the <code>Linux for Tegra</code> directory and run <code>sudo ./nvsdkmanager_flash.sh</code>. When prompted, disconnect the Jetson from host device and allow it to boot. Congratulations, you have successfully flashed your Jetson with custom firmware. 0a866bc6bfccea12e15b2b9c202120e3f41e03ff 1830 1829 2024-08-30T19:55:57Z Vedant 24 /* Building Custom Kernel and Installing Modules */ wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: <code>https://developer.nvidia.com/sdk-manager</code> # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the <code>kernel</code> eventually navigate to the <code>kernel-jammy-src</code> folder and run <code>make menuconfig ARCH=arm64</code>. This will bring up a UI with configurable drivers and peripherals. Select desired configurations and save. # The configurations can be found within a <code>.config</code> file located within the same directory. Copy the contents and locate the <code>defconfig</code> file in <code>./arch/arm64/configs/</code>, overwriting it with the copied contents. == Building Custom Kernel and Installing Modules == # Navigate back out to <code>sources</code>. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE=<toolchain-path>/bin/aarch64-buildroot-linux-gnu-</code>. If installation was done correctly as per the pre-requisites section, the command <code>export CROSS_COMPILE=$HOME/l4t-gcc/aarch64--glibc--stable-2022.08-1/bin/aarch64-buildroot-linux-gnu-</code> should work. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE_AARCH64_PATH=</code>, and <code>export CROSS_COMPILE_AARCH64=/bin/aarch64-buildroot-linux-gnu-</code>. (Potentially deprecated) # Inside the sources, directory, make an output directory for built kernel files using <code>mkdir kernel_out</code>. # Build the modules using the command <code>./nvbuild.sh -o kernel_out</code>. This will compile the drivers and device trees for the new kernel. # Navigate out from the <code>sources</code> directory into the <code>Linux for Tegra</code>. # Use the <code>cp</code> to overwrite <code>./rootfs/usr/lib/modules/5.15.136-tegra/updates/nvgpu.ko</code> with <code>./source/kernel_out/nvgpu/drivers/gpu/nvgpu/nvgpu.ko</code>. # Repeat the previous step to replace <code>Linux_for_Tegra/kernel/dtb/</code> with </code>source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/dts/nvidia</code>. Ensure that instead of overwriting the directory, only the files are copied over. # Overwrite the Image file in <code>./kernel</code> with <code>./source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/Image </code>. # To specify the installation path for the compiled modules, use the command <code>export INSTALL_MOD_PATH=$HOME/nvidia/nvidia_sdk/JetPack_6.0_Linux_JETSON_AGX_ORIN_TARGETS/Linux_for_Tegra/rootfs/</code>. # Install modules using the command <code>./nvbuild -i</code>. The Jetson is now ready to be flashed. == Flashing the Kernel == Ensure that the target Jetson is connected to the host device and is in recovery mode. Navigate to the <code>Linux for Tegra</code> directory and run <code>sudo ./nvsdkmanager_flash.sh</code>. When prompted, disconnect the Jetson from host device and allow it to boot. Congratulations, you have successfully flashed your Jetson with custom firmware. 8c848c263b78b63b0e05a2211cc99ba3f2f27518 1831 1830 2024-08-30T19:56:24Z Vedant 24 /* Flashing the Kernel */ wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: <code>https://developer.nvidia.com/sdk-manager</code> # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the <code>kernel</code> eventually navigate to the <code>kernel-jammy-src</code> folder and run <code>make menuconfig ARCH=arm64</code>. This will bring up a UI with configurable drivers and peripherals. Select desired configurations and save. # The configurations can be found within a <code>.config</code> file located within the same directory. Copy the contents and locate the <code>defconfig</code> file in <code>./arch/arm64/configs/</code>, overwriting it with the copied contents. == Building Custom Kernel and Installing Modules == # Navigate back out to <code>sources</code>. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE=<toolchain-path>/bin/aarch64-buildroot-linux-gnu-</code>. If installation was done correctly as per the pre-requisites section, the command <code>export CROSS_COMPILE=$HOME/l4t-gcc/aarch64--glibc--stable-2022.08-1/bin/aarch64-buildroot-linux-gnu-</code> should work. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE_AARCH64_PATH=</code>, and <code>export CROSS_COMPILE_AARCH64=/bin/aarch64-buildroot-linux-gnu-</code>. (Potentially deprecated) # Inside the sources, directory, make an output directory for built kernel files using <code>mkdir kernel_out</code>. # Build the modules using the command <code>./nvbuild.sh -o kernel_out</code>. This will compile the drivers and device trees for the new kernel. # Navigate out from the <code>sources</code> directory into the <code>Linux for Tegra</code>. # Use the <code>cp</code> to overwrite <code>./rootfs/usr/lib/modules/5.15.136-tegra/updates/nvgpu.ko</code> with <code>./source/kernel_out/nvgpu/drivers/gpu/nvgpu/nvgpu.ko</code>. # Repeat the previous step to replace <code>Linux_for_Tegra/kernel/dtb/</code> with </code>source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/dts/nvidia</code>. Ensure that instead of overwriting the directory, only the files are copied over. # Overwrite the Image file in <code>./kernel</code> with <code>./source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/Image </code>. # To specify the installation path for the compiled modules, use the command <code>export INSTALL_MOD_PATH=$HOME/nvidia/nvidia_sdk/JetPack_6.0_Linux_JETSON_AGX_ORIN_TARGETS/Linux_for_Tegra/rootfs/</code>. # Install modules using the command <code>./nvbuild -i</code>. The Jetson is now ready to be flashed. == Flashing the Kernel == Ensure that the target Jetson is connected to the host device and is in recovery mode. Navigate to the <code>Linux for Tegra</code> directory and run <code>sudo ./nvsdkmanager_flash.sh</code>. When prompted, disconnect the Jetson from host device and allow it to boot. Congratulations, you have successfully flashed your Jetson with custom firmware. 2f886fe526ac33b1b2ce7060454ea3613933b307 1832 1831 2024-08-30T19:56:36Z Vedant 24 /* Flashing the Kernel */ wikitext text/x-wiki = Flashing Standard Firmware = == SDKManager == SDKManager is available only on Linux, and can be installed here: <code>https://developer.nvidia.com/sdk-manager</code> # Start up the SDKManager # Put the Jetson into recovery mode. For the AGX, this can be done by pressing the recovery button while powering on the device. For the Nano and NX, however, a jumper will be required. # Connect the Target Jetson to the host device and ensure that the target device is recognized. # Follow the instructions on the application, choosing configurations as necessary. = Flashing Custom Firmware (For Jetson 36.3) = == Pre-requisites == # Please install required packages with the command <code>sudo apt install build-essential bc && sudo apt install build-essential bc</code>. === Downloading the Toolchain === # Download the Toolchain binaries located in <code>https://developer.nvidia.com/embedded/jetson-linux</code>. # From there, <code>mkdir $HOME/l4t-gcc</code>, <code>cd $HOME/l4t-gcc</code> and extract the installed toolchain into this newly created directory using the <code>tar</code> command. == Downloading the Kernel == # Follow steps 1 to 3 as mentioned in Flashing Standard Firmware. # Proceed to the second step of the SDKManager, where the respective individual dependencies and Jetson Images are listed and are to be installed. Proceed with the installation. # When prompted to actually flash the Jetson, opt to skip. This will install the <code>nvidia</code> folder on your home directory, in which the <code>rootfs</code>, <code>kernel</code>, and <code>bootloader</code> are located. # Navigate to <code>nvidia</code> and <code>cd</code> through its subdirectories, until <code>Linux for Tegra</code> is reached. # Inside <code>Linux for Tegra</code>, <code>cd</code> into the <code>sources</code> folder. It should be unpopulated with the exception of some bash scripts. Run the <code>source_sync.sh</code> script and when asked to specify the release tag of the downloadable sources, enter <code>jetson_36.3</code>. This will install the sources for the respective Jetson version as necessary. To find the release tag of future iterations of the Jetson firmware, please refer to its respective release notes. # Once sources have been synced, the <code>sources</code> directory should now be populated with the required files. == Customizing Kernel == # Within <code>source</code>, enter the <code>kernel</code> eventually navigate to the <code>kernel-jammy-src</code> folder and run <code>make menuconfig ARCH=arm64</code>. This will bring up a UI with configurable drivers and peripherals. Select desired configurations and save. # The configurations can be found within a <code>.config</code> file located within the same directory. Copy the contents and locate the <code>defconfig</code> file in <code>./arch/arm64/configs/</code>, overwriting it with the copied contents. == Building Custom Kernel and Installing Modules == # Navigate back out to <code>sources</code>. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE=<toolchain-path>/bin/aarch64-buildroot-linux-gnu-</code>. If installation was done correctly as per the pre-requisites section, the command <code>export CROSS_COMPILE=$HOME/l4t-gcc/aarch64--glibc--stable-2022.08-1/bin/aarch64-buildroot-linux-gnu-</code> should work. # Define the Cross-compilation toolchain with the commands <code>export CROSS_COMPILE_AARCH64_PATH=</code>, and <code>export CROSS_COMPILE_AARCH64=/bin/aarch64-buildroot-linux-gnu-</code>. (Potentially deprecated) # Inside the sources, directory, make an output directory for built kernel files using <code>mkdir kernel_out</code>. # Build the modules using the command <code>./nvbuild.sh -o kernel_out</code>. This will compile the drivers and device trees for the new kernel. # Navigate out from the <code>sources</code> directory into the <code>Linux for Tegra</code>. # Use the <code>cp</code> to overwrite <code>./rootfs/usr/lib/modules/5.15.136-tegra/updates/nvgpu.ko</code> with <code>./source/kernel_out/nvgpu/drivers/gpu/nvgpu/nvgpu.ko</code>. # Repeat the previous step to replace <code>Linux_for_Tegra/kernel/dtb/</code> with </code>source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/dts/nvidia</code>. Ensure that instead of overwriting the directory, only the files are copied over. # Overwrite the Image file in <code>./kernel</code> with <code>./source/kernel_out/kernel/kernel-jammy-src/arch/arm64/boot/Image </code>. # To specify the installation path for the compiled modules, use the command <code>export INSTALL_MOD_PATH=$HOME/nvidia/nvidia_sdk/JetPack_6.0_Linux_JETSON_AGX_ORIN_TARGETS/Linux_for_Tegra/rootfs/</code>. # Install modules using the command <code>./nvbuild -i</code>. The Jetson is now ready to be flashed. == Flashing the Kernel == Ensure that the target Jetson is connected to the host device and is in recovery mode. Navigate to the <code>Linux for Tegra</code> directory and run <code>sudo ./nvsdkmanager_flash.sh</code>. When prompted, disconnect the Jetson from host device and allow it to boot. Congratulations, you have successfully flashed your Jetson with custom firmware. 8c848c263b78b63b0e05a2211cc99ba3f2f27518 0 218 1824 1441 2024-08-29T04:59:11Z Budzianowski 19 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run <code>sudo ./tools/l4t_flash_prerequisites.sh</code> * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> under NVIDIA CORP. APX * Run <code>./flash.sh </code> Different options for different usecases(https://docs.nvidia.com/jetson/archives/r36.3/DeveloperGuide/IN/QuickStart.html#in-quickstart) Jetson AGX Orin Developer Kit (eMMC): <code>$ sudo ./flash.sh jetson-agx-orin-devkit internal</code> * Watch for few minutes, typically it crashes early, then go for lunch. === Upgrading Nano to Jetson Linux 36.3 === ==== Buildroot approach (avoids SDK Manager) ==== sudo mkfs.ext4 /dev/sdb sudo umount /dev/sdb lsblk -f * Transfer rootfs onto sd card sudo dd if=/home/dpsh/Downloads/rootfs.ext4 of=/dev/sdc1 bs=1M status=progress * Make sure the data was transferred to SD/NVMe sync [[Category: Firmware]] 8e242c09b16592da87cabe1bd4ceec9d04ca3329 6 421 1833 2024-09-01T01:28:50Z Goblinrum 25 wikitext text/x-wiki BCM Pinout names for Jetson 99ccf6f68f4e2fd72ca8353d17c4471f5c21a831 0 422 1834 2024-09-01T01:55:56Z Goblinrum 25 Created page with "Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note th..." wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the 1.5 inch LCD Module guide. # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |} # Clone the repository and run the code. You may need to install the following dependencies: TBD 75f55547dc83d7e670fd48303a13509f946e8192 1835 1834 2024-09-01T01:56:38Z Goblinrum 25 wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the 1.5 inch LCD Module guide. # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |- | Example || Example || Example |} # Clone the repository and run the code. You may need to install the following dependencies: TBD [[Category:Electronics]] [[Category:Hardware]] 0d010861374cc127382bf05d9b1fff01bef2e185 1836 1835 2024-09-01T02:07:25Z Goblinrum 25 wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the 1.5 inch LCD Module guide. # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | 3v3 || 1 || 3v3 |- | GND || 6 || GND |- | DIN/MOSI || 19 || D10 (Automatically selected by Jetson-GPIO) |- | SCK || 23 || D11 (Automatically selected) |- | CS || 24 || CE0 (Automatically selected) |- | DC || 22 || D25 |- | RST || 31 || D6 |- | BL || 32 || D12 |} # Clone the repository and run the code. You may need to install the following dependencies: TBD [[Category:Electronics]] [[Category:Hardware]] eb7c3c777a5bf2f1ac8bff532d8bc49069d4054d 1837 1836 2024-09-01T02:12:21Z Goblinrum 25 wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the 1.5 inch LCD Module guide. # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | 3v3 || 1 || 3v3 |- | GND || 6 || GND |- | DIN/MOSI || 19 || D10 (Automatically selected by Jetson-GPIO) |- | SCK || 23 || D11 (Automatically selected) |- | CS || 24 || CE0 (Automatically selected) |- | DC || 22 || D25 |- | RST || 31 || D6 |- | BL || 32 || D12 |} Clone the repository and run the code. You may need to install the following pip dependencies: * Jetson.GPIO * adafruit-circuitpython-busdevice * spidev * adafruit-blinka-displayio * adafruit-circuitpython-rgb-display * adafruit-circuitpython-st7789 * Pillow [[Category:Electronics]] [[Category:Hardware]] 9c0a1a40ccd459a7e5ba7ec9f76ca4a5d6b5a743 0 422 1838 1837 2024-09-01T02:13:24Z Goblinrum 25 wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the 1.5 inch LCD Module guide. # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | 3v3 || 1 || 3v3 |- | GND || 6 || GND |- | DIN/MOSI || 19 || D10 (Automatically selected by Jetson-GPIO) |- | SCK || 23 || D11 (Automatically selected) |- | CS || 24 || CE0 (Automatically selected) |- | DC || 22 || D25 |- | RST || 31 || D6 |- | BL || 32 || D12 |} Clone the repository and launch run.py. You may need to install the following pip dependencies: * Jetson.GPIO * adafruit-circuitpython-busdevice * spidev * adafruit-blinka-displayio * adafruit-circuitpython-rgb-display * adafruit-circuitpython-st7789 * Pillow The example code should draw some shapes every 2 seconds, followed by the Waveshare logo, then a Yuan dynasty poem, and lastly a series of images. [[Category:Electronics]] [[Category:Hardware]] cb571f95e5c592772ce3e0c3ca12dc1ced632b3e 1839 1838 2024-09-01T02:18:47Z Goblinrum 25 wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the [https://www.waveshare.com/wiki/1.5inch_LCD_Module#Jetson_Nano_User_Guide Jetson Nano section in the 1.5in LCD Guide] # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | 3v3 || 1 || 3v3 |- | GND || 6 || GND |- | DIN/MOSI || 19 || D10 (Automatically selected by Jetson-GPIO) |- | SCK || 23 || D11 (Automatically selected) |- | CS || 24 || CE0 (Automatically selected) |- | DC || 22 || D25 |- | RST || 31 || D6 |- | BL || 32 || D12 |} Clone the repository and launch run.py. You may need to install the following pip dependencies: * Jetson.GPIO * adafruit-circuitpython-busdevice * spidev * adafruit-blinka-displayio * adafruit-circuitpython-rgb-display * adafruit-circuitpython-st7789 * Pillow The example code should draw some shapes every 2 seconds, followed by the Waveshare logo, then a Yuan dynasty poem, and lastly a series of images. [[Category:Electronics]] [[Category:Hardware]] 3883a329872d7bb587a41ff9ebe452eb6ae2fa8e 1840 1839 2024-09-01T02:19:32Z Goblinrum 25 wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the [https://www.waveshare.com/wiki/1.5inch_LCD_Module#Jetson_Nano_User_Guide Jetson Nano section in the 1.5in LCD Guide]. Follow the "Enable SPI" and "Library Installation" sections. Stop before the "Download Example" section. # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | 3v3 || 1 || 3v3 |- | GND || 6 || GND |- | DIN/MOSI || 19 || D10 (Automatically selected by Jetson-GPIO) |- | SCK || 23 || D11 (Automatically selected) |- | CS || 24 || CE0 (Automatically selected) |- | DC || 22 || D25 |- | RST || 31 || D6 |- | BL || 32 || D12 |} Clone the repository and launch run.py. You may need to install the following pip dependencies: * Jetson.GPIO * adafruit-circuitpython-busdevice * spidev * adafruit-blinka-displayio * adafruit-circuitpython-rgb-display * adafruit-circuitpython-st7789 * Pillow The example code should draw some shapes every 2 seconds, followed by the Waveshare logo, then a Yuan dynasty poem, and lastly a series of images. [[Category:Electronics]] [[Category:Hardware]] c79acdb372a5c077bd10dde5cb3753ac8734d6b0 1841 1840 2024-09-01T23:58:59Z Goblinrum 25 wikitext text/x-wiki Instructions to run the [https://www.waveshare.com/wiki/1.69inch_LCD_Module Waveshare 1.69in LCDs] on the Jetson carrier boards [[File:Jetson BCM Mapping.png|thumb]] Note the Linux BCM mappings in the image shown. Theoretically with the Jetson.GPIO library, you are able to use either the physical board pin mapping or the BCM mapping, but this process has only been vetted for the BCM mode. When referring to pins in the Python package, use the BCM mapping. For example, if you use pin 32, use "D12" when referring to the pin. With the Adafruit package, this is written as "board.D12" The [https://www.waveshare.com/wiki/1.5inch_LCD_Module 1.5 Inch LCD Module] page actually tells you how to set up the Jetson for the Waveshare LCDs. However, since it uses a different controller, the actual firmware needs to be modified. You can find the draft changes repo [https://github.com/goblinrum/waveshare-1.69in-lcd-jetson here]. Steps to reproduce: # On Jetson, use the jetson-io to enable SPI1. Steps are shown in the [https://www.waveshare.com/wiki/1.5inch_LCD_Module#Jetson_Nano_User_Guide Jetson Nano section in the 1.5in LCD Guide]. Follow the "Enable SPI" and "Library Installation" sections. Stop before the "Download Example" section. # Power down the Jetson and wire up the display using the following pinout: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | 3v3 || 1 || 3v3 |- | GND || 6 || GND |- | DIN/MOSI || 19 || D10 (Automatically selected by Jetson-GPIO) |- | SCK || 23 || D11 (Automatically selected) |- | CS || 24 || CE0 (Automatically selected) |- | DC || 22 || D25 |- | RST || 31 || D6 |- | BL || 32 || D12 |} Clone the repository and launch run.py. You may need to install the following pip dependencies: * Jetson.GPIO * adafruit-circuitpython-busdevice * spidev * adafruit-blinka-displayio * adafruit-circuitpython-rgb-display * adafruit-circuitpython-st7789 * Pillow You may also need to run "sudo modprobe spidev" to get the spidev driver loaded. The example code should draw some shapes every 2 seconds, followed by the Waveshare logo, then a Yuan dynasty poem, and lastly a series of images. == Testing with 2 LCDs == Ideally, we run both screens off the same SPI line and use the built-in chip-select pins on the Jetson. However, based on current testing, the Jetson-IO CS functionality does not work. If a device is connected to CS1 and CS0, CS0 will always remain low. The code in the repository assumes you use both spi lines (SPI0 and SPI2, so spidev0.0 and spidev2.0 respectively). Wire up the second display as such: {| class="wikitable" |- ! Function !! Jetson Pin !! BCM Pin |- | 3v3 || 17 || 3v3 |- | GND || 9 || GND |- | DIN/MOSI || 37 || D26 (Automatically selected by Jetson-GPIO) |- | SCK || 13 || D27 (Automatically selected) |- | CS || 18 || D24 (Automatically selected) |- | DC || 36 || D16 |- | RST || 38 || D20 |- | BL || 33 || D13 |} The code in the repository should already specify these pins for the second screen. [[Category:Electronics]] [[Category:Hardware]] 37a0d6cb66a85e591d325ab077ebee376e09205d 0 218 1842 1824 2024-09-05T16:08:22Z Budzianowski 19 /* Buildroot approach (avoids SDK Manager) */ wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run <code>sudo ./tools/l4t_flash_prerequisites.sh</code> * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> under NVIDIA CORP. APX * Run <code>./flash.sh </code> Different options for different usecases(https://docs.nvidia.com/jetson/archives/r36.3/DeveloperGuide/IN/QuickStart.html#in-quickstart) Jetson AGX Orin Developer Kit (eMMC): <code>$ sudo ./flash.sh jetson-agx-orin-devkit internal</code> * Watch for few minutes, typically it crashes early, then go for lunch. === Upgrading Nano to Jetson Linux 36.3 === ==== Buildroot approach (avoids SDK Manager) ==== * sudo mkfs.ext4 /dev/sdb * sudo umount /dev/sdb * lsblk -f * Transfer rootfs onto sd card sudo dd if=/home/dpsh/Downloads/rootfs.ext4 of=/dev/sdc1 bs=1M status=progress * Make sure the data was transferred to SD/NVMe sync [[Category: Firmware]] e860a6679ea564200e493b82b88d39453461a277 1843 1842 2024-09-05T16:09:00Z Budzianowski 19 wikitext text/x-wiki Notes on programming/interfacing with Jetson Orin hardware. === Upgrading AGX to Jetson Linux 36.3 === ==== BSP approach (avoids SDK Manager) ==== * Requires Ubuntu 22.04. Very unhappy to work on Gentoo. * Requires Intel/AMD 64bit CPU. * Download "Driver Package (BSP)" from [https://developer.nvidia.com/embedded/jetson-linux here] * Unpack (as root, get used to doing most of this as root), preserving privileges ** <code>tar xjpf ...</code> * Download "Sample Root Filesystem" * Unpack (as root..) into rootfs directory inside of the BSP archive above. * Run <code>sudo ./tools/l4t_flash_prerequisites.sh</code> * Run <code>./apply_binaries.sh</code> from the BSP ** Note: If apply_binaries (or frankly, anything, this is brittle) fails, remove and recreate rootfs - the OS might be left in an unbootable state. * Reboot AGX into "Recovery Mode" - hold the recovery button and reset button, release simultaneously ((sic) reset first?) * Connect USB-C cable to the debug port ("front" USB-c) * Nvidia AGX device should appear in the <code>lsusb</code> under NVIDIA CORP. APX * Run <code>./flash.sh </code> Different options for different usecases(https://docs.nvidia.com/jetson/archives/r36.3/DeveloperGuide/IN/QuickStart.html#in-quickstart) Jetson AGX Orin Developer Kit (eMMC): <code>$ sudo ./flash.sh jetson-agx-orin-devkit internal</code> * Watch for few minutes, typically it crashes early, then go for lunch. === Upgrading Nano to Jetson Linux 36.3 === ==== Buildroot approach (avoids SDK Manager) ==== * <code>sudo mkfs.ext4 /dev/sdb </code> * <code>sudo umount /dev/sdb </code> * <code>lsblk -f</code> * Transfer rootfs onto sd card <code>sudo dd if=/home/dpsh/Downloads/rootfs.ext4 of=/dev/sdc1 bs=1M status=progress</code> * Make sure the data was transferred to SD/NVMe <code>sync</code> [[Category: Firmware]] ed5e5c1f2ecdb9ea931e6d41780bafa988339127 0 293 1844 1331 2024-09-18T02:36:24Z Ben 2 /* Papers */ wikitext text/x-wiki === Papers === * [https://distill.pub/2017/ctc/ CTC] * [https://arxiv.org/abs/1810.04805 BERT] * [https://arxiv.org/abs/2006.11477 wav2vec 2.0] * [https://arxiv.org/abs/2106.07447 HuBERT] * [https://speechbot.github.io/ Textless NLP project] * [https://arxiv.org/abs/2210.13438 Encodec] * [https://arxiv.org/abs/2308.16692 SpeechTokenizer] * [https://github.com/suno-ai/bark Suno Bark Model] * [https://github.com/google-research/parti Parti] f34bd46589e075ee29329d1e397f2bdb29fbd34b 0 425 1847 2024-09-25T00:29:18Z Vrtnis 21 /*Initial Commit*/ wikitext text/x-wiki {| class="wikitable" ! Session Code !! Session Title !! Session Type !! Paper Code !! Paper Title !! Authors !! Affiliations |- | ThBT12 || Learning from Humans || Regular session || ThBT12.3 || Learning Human-To-Humanoid Real-Time Whole-Body Teleoperation || He, Tairan; Luo, Zhengyi; Xiao, Wenli; Zhang, Chong; Kitani, Kris; Liu, Changliu; Shi, Guanya || Carnegie Mellon University; ETH Zurich |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.3 || Whole-Body Humanoid Robot Locomotion with Human Reference || Zhang, Qiang; Cui, Peter; Yan, David; Sun, Jingkai; Duan, Yiqun; Han, Gang; Zhao, Wen; Zhang, Weining; Guo, Yijie; Zhang, Arthur; Xu, Renjing || The Hong Kong University of Science and Technology; Peter & David Robotics; University of Technology Sydney; PND Robotics; UBTECH Robotics |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.13 || Fly by Book: How to Train a Humanoid Robot to Fly an Airplane Using Large Language Models || Kim, Hyungjoo; Min, Sungjae; Kang, Gyuree; Kim, Jihyeok; Shim, David Hyunchul || KAIST |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.14 || Towards Designing a Low-Cost Humanoid Robot with Flex Sensors-Based Movement || Al Omoush, Muhammad H.; Kishore, Sameer; Mehigan, Tracey || Dublin City University; Middlesex University |- | MoWMT17 || AI Meets Autonomy: Vision, Language, and Autonomous Systems || Workshop || MoWMT17.1 || AI Meets Autonomy: Vision, Language, and Autonomous Systems || Wang, Wenshan; Zhang, Ji; Zhang, Haochen; Zhao, Shibo; Gupta, Abhinav; Ramanan, Deva; Zeng, Andy; Kim, Ayoung; Nieto-Granda, Carlos || Carnegie Mellon University; Google DeepMind; Seoul National University; DEVCOM U.S. Army Research Laboratory |- | MoWMT18 || Collecting, Managing and Utilizing Data through Embodied Robots || Workshop || MoWMT18.1 || Collecting, Managing and Utilizing Data through Embodied Robots || Saito, Namiko; Al-Sada, Mohammed; Shigemune, Hiroki; Tsumura, Ryosuke; Funabashi, Satoshi; Miyake, Tamon; Ogata, Tetsuya || The University of Edinburgh; Waseda University, Qatar University; Shibaura Institute of Technology; AIST; Waseda University |- | MoWPT2 || Long-Term Perception for Autonomy in Dynamic Human-Centric Environments: What Do Robots Need? || Workshop || MoWPT2.1 || Long-Term Perception for Autonomy in Dynamic Human-Centric Environments: What Do Robots Need? || Schmid, Lukas M.; Talak, Rajat; Zheng, Jianhao; Andersson, Olov; Oleynikova, Helen; Park, Jong Jin; Wald, Johanna; Siegwart, Roland; Tombari, Federico; Carlone, Luca || MIT; Stanford University; KTH Royal Institute; ETH Zurich; Amazon Lab126; Everyday Robots; Technische Universität München |- | TuWAT5 || Humanoid Hybrid Sprint || Tutorial || TuWAT5.1 || Humanoid Hybrid Sprint || Osokin, Ilya || Moscow Institute of Physics and Technology |- | TuBT11 || Legged and Humanoid Robots || Regular session || TuBT11.1 || Invariant Smoother for Legged Robot State Estimation with Dynamic Contact Event Information (I) || Yoon, Ziwon; Kim, Joon-Ha; Park, Hae-Won || Georgia Institute of Technology; KAIST |- | TuBT11 || Legged and Humanoid Robots || Regular session || TuBT11.2 || MorAL: Learning Morphologically Adaptive Locomotion Controller for Quadrupedal Robots on Challenging Terrains || Luo, Zeren; Dong, Yinzhao; Li, Xinqi; Huang, Rui; Shu, Zhengjie; Xiao, Erdong; Lu, Peng || The University of Hong Kong |- | WeAT1 || Embodied AI with Two Arms: Zero-Shot Learning, Safety and Modularity || Regular session || WeAT1.4 || Embodied AI with Two Arms: Zero-Shot Learning, Safety and Modularity || Varley, Jacob; Singh, Sumeet; Jain, Deepali; Choromanski, Krzysztof; Zeng, Andy; Basu Roy Chowdhury, Somnath; Dubey, Avinava; Sindhwani, Vikas || Google; Google DeepMind; UNC Chapel Hill; Google Brain, NYC |- | WeDT12 || Imitation Learning I || Regular session || WeDT12.1 || Uncertainty-Aware Haptic Shared Control with Humanoid Robots for Flexible Object Manipulation || Hara, Takumi; Sato, Takashi; Ogata, Tetsuya; Awano, Hiromitsu || Kyoto University; Waseda University |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.16 || From CAD to URDF: Co-Design of a Jet-Powered Humanoid Robot Including CAD Geometry || Vanteddu, Punith Reddy; Nava, Gabriele; Bergonti, Fabio; L'Erario, Giuseppe; Paolino, Antonello; Pucci, Daniele || Istituto Italiano Di Tecnologia |} 79687b1092bae56ead9433ab3d7a8fb333a927fc 1848 1847 2024-09-25T00:30:36Z Vrtnis 21 Vrtnis moved page [[IROS 24 Humanoid]] to [[IROS 24 Humanoid Papers]] wikitext text/x-wiki {| class="wikitable" ! Session Code !! Session Title !! Session Type !! Paper Code !! Paper Title !! Authors !! Affiliations |- | ThBT12 || Learning from Humans || Regular session || ThBT12.3 || Learning Human-To-Humanoid Real-Time Whole-Body Teleoperation || He, Tairan; Luo, Zhengyi; Xiao, Wenli; Zhang, Chong; Kitani, Kris; Liu, Changliu; Shi, Guanya || Carnegie Mellon University; ETH Zurich |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.3 || Whole-Body Humanoid Robot Locomotion with Human Reference || Zhang, Qiang; Cui, Peter; Yan, David; Sun, Jingkai; Duan, Yiqun; Han, Gang; Zhao, Wen; Zhang, Weining; Guo, Yijie; Zhang, Arthur; Xu, Renjing || The Hong Kong University of Science and Technology; Peter & David Robotics; University of Technology Sydney; PND Robotics; UBTECH Robotics |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.13 || Fly by Book: How to Train a Humanoid Robot to Fly an Airplane Using Large Language Models || Kim, Hyungjoo; Min, Sungjae; Kang, Gyuree; Kim, Jihyeok; Shim, David Hyunchul || KAIST |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.14 || Towards Designing a Low-Cost Humanoid Robot with Flex Sensors-Based Movement || Al Omoush, Muhammad H.; Kishore, Sameer; Mehigan, Tracey || Dublin City University; Middlesex University |- | MoWMT17 || AI Meets Autonomy: Vision, Language, and Autonomous Systems || Workshop || MoWMT17.1 || AI Meets Autonomy: Vision, Language, and Autonomous Systems || Wang, Wenshan; Zhang, Ji; Zhang, Haochen; Zhao, Shibo; Gupta, Abhinav; Ramanan, Deva; Zeng, Andy; Kim, Ayoung; Nieto-Granda, Carlos || Carnegie Mellon University; Google DeepMind; Seoul National University; DEVCOM U.S. Army Research Laboratory |- | MoWMT18 || Collecting, Managing and Utilizing Data through Embodied Robots || Workshop || MoWMT18.1 || Collecting, Managing and Utilizing Data through Embodied Robots || Saito, Namiko; Al-Sada, Mohammed; Shigemune, Hiroki; Tsumura, Ryosuke; Funabashi, Satoshi; Miyake, Tamon; Ogata, Tetsuya || The University of Edinburgh; Waseda University, Qatar University; Shibaura Institute of Technology; AIST; Waseda University |- | MoWPT2 || Long-Term Perception for Autonomy in Dynamic Human-Centric Environments: What Do Robots Need? || Workshop || MoWPT2.1 || Long-Term Perception for Autonomy in Dynamic Human-Centric Environments: What Do Robots Need? || Schmid, Lukas M.; Talak, Rajat; Zheng, Jianhao; Andersson, Olov; Oleynikova, Helen; Park, Jong Jin; Wald, Johanna; Siegwart, Roland; Tombari, Federico; Carlone, Luca || MIT; Stanford University; KTH Royal Institute; ETH Zurich; Amazon Lab126; Everyday Robots; Technische Universität München |- | TuWAT5 || Humanoid Hybrid Sprint || Tutorial || TuWAT5.1 || Humanoid Hybrid Sprint || Osokin, Ilya || Moscow Institute of Physics and Technology |- | TuBT11 || Legged and Humanoid Robots || Regular session || TuBT11.1 || Invariant Smoother for Legged Robot State Estimation with Dynamic Contact Event Information (I) || Yoon, Ziwon; Kim, Joon-Ha; Park, Hae-Won || Georgia Institute of Technology; KAIST |- | TuBT11 || Legged and Humanoid Robots || Regular session || TuBT11.2 || MorAL: Learning Morphologically Adaptive Locomotion Controller for Quadrupedal Robots on Challenging Terrains || Luo, Zeren; Dong, Yinzhao; Li, Xinqi; Huang, Rui; Shu, Zhengjie; Xiao, Erdong; Lu, Peng || The University of Hong Kong |- | WeAT1 || Embodied AI with Two Arms: Zero-Shot Learning, Safety and Modularity || Regular session || WeAT1.4 || Embodied AI with Two Arms: Zero-Shot Learning, Safety and Modularity || Varley, Jacob; Singh, Sumeet; Jain, Deepali; Choromanski, Krzysztof; Zeng, Andy; Basu Roy Chowdhury, Somnath; Dubey, Avinava; Sindhwani, Vikas || Google; Google DeepMind; UNC Chapel Hill; Google Brain, NYC |- | WeDT12 || Imitation Learning I || Regular session || WeDT12.1 || Uncertainty-Aware Haptic Shared Control with Humanoid Robots for Flexible Object Manipulation || Hara, Takumi; Sato, Takashi; Ogata, Tetsuya; Awano, Hiromitsu || Kyoto University; Waseda University |- | FrPI6T1 || Humanoid and Bipedal Locomotion || Teaser Session || FrPI6T1.16 || From CAD to URDF: Co-Design of a Jet-Powered Humanoid Robot Including CAD Geometry || Vanteddu, Punith Reddy; Nava, Gabriele; Bergonti, Fabio; L'Erario, Giuseppe; Paolino, Antonello; Pucci, Daniele || Istituto Italiano Di Tecnologia |} 79687b1092bae56ead9433ab3d7a8fb333a927fc 0 426 1849 2024-09-25T00:30:36Z Vrtnis 21 Vrtnis moved page [[IROS 24 Humanoid]] to [[IROS 24 Humanoid Papers]] wikitext text/x-wiki #REDIRECT [[IROS 24 Humanoid Papers]] fcc971bed200572db45aba48064485d1a045ca43 0 1 1850 1825 2024-09-25T03:01:07Z Ben 2 add moteus n1 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] 2dabce6cd0bcc0c155449d9df615944b68aa0a92 1860 1850 2024-10-15T00:44:11Z Ben 2 /* Actuators */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] a8dc4f1b6b1d4d006713d09bb23de4059b9a1485 1875 1860 2024-10-15T20:10:16Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} === Inference Boards === {| class="wikitable" |- ! [[Jetson]] |- ! [[SiMi AI]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] 5a7b9ab145828ab77f2e8b72ef7ae560d8a96943 1878 1875 2024-10-15T20:11:50Z Ben 2 /* Inference Boards */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} === Inference Boards === {| class="wikitable" ! Inference Boards |- ! [[Jetson]] |- ! [[SiMi AI]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] 465ac1454637d1479f82240e48386ceea17fecc0 1879 1878 2024-10-15T20:12:02Z Ben 2 /* Inference Boards */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Actuators === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} === Inference Boards === {| class="wikitable" ! Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} === Motor Controllers === {| class="wikitable" |- ! Controller |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] d7232270c2f4dc8d43764e0a27f63916b48efda3 1880 1879 2024-10-15T20:12:42Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable" ! Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable" |- ! Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] 0818baec87ebf7beb8739466be9a07c361eb2d47 1881 1880 2024-10-15T20:13:10Z Ben 2 /* Components */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable" ! Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable" |- ! Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable" |- ! Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] 14ba2d4d342c28fa83de1ef3bb6b5fcbb8f3bf6d 1882 1881 2024-10-15T20:13:49Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community]! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable" ! Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable" |- ! Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable" |- ! Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Discord community === [https://discord.gg/kscale Discord] 5a7e5ca3afdbb2309d27b7824752d6d36a9126da 1883 1882 2024-10-15T20:14:08Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable" |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable" ! Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable" |- ! Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable" |- ! Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} 2b5f4e9d54e910831c5f7167cf38ab9a475912fe 1884 1883 2024-10-15T20:17:43Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable mw-collapsible mw-collapsed" |+ Resources |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable" |- ! Actuator |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable" ! Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable" |- ! Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable" |- ! Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} d83f502a72e2747166766e3006da305da416f76d 1885 1884 2024-10-15T20:23:47Z Ben 2 /* Components */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable mw-collapsible mw-collapsed" |+ Resources |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable mw-collapsible mw-collapsed" |+ Actuators |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} === Communication Protocols === {| class="wikitable" |- ! Name |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} 9e4c5e2c0f7635ec53870629de045b8f1617d0af 1886 1885 2024-10-15T20:24:01Z Ben 2 wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable mw-collapsible mw-collapsed" |+ Resources |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable mw-collapsible mw-collapsed" |+ Actuators |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[Agibot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} 8d07cc75401379baafe2d4995a4ac0c15bce6621 1887 1886 2024-10-24T20:37:44Z Ben 2 /* Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable mw-collapsible mw-collapsed" |+ Resources |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable mw-collapsible mw-collapsed" |+ Actuators |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[AGIBot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} 875143050a010a74e17a70cca80b02d329373bf5 1895 1887 2024-11-01T09:44:13Z Kris 97 /* Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable mw-collapsible mw-collapsed" |+ Resources |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable mw-collapsible mw-collapsed" |+ Actuators |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[AGIBot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Humanoid]] | |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} 921072b715bb7ea3039c3bfdc708768db9b64af4 0 427 1851 2024-09-25T03:02:48Z Ben 2 Created page with "The [https://mjbots.com/products/moteus-n1 moteus-n1] is a motor controller from MJBots. The project is open-source, with firmware available [https://github.com/mjbots/moteus..." wikitext text/x-wiki The [https://mjbots.com/products/moteus-n1 moteus-n1] is a motor controller from MJBots. The project is open-source, with firmware available [https://github.com/mjbots/moteus here]. {{infobox actuator |name = moteus-n1 |manufacturer = MJBots |cost = |purchase_link = https://mjbots.com/products/moteus-n1 |nominal_torque = |peak_torque = |weight = |dimensions = |gear_ratio = |voltage = 10 to 54VDC |cad_link = https://mjbots.com/products/moteus-n1 |interface = CAN-FD |gear_type = }} 3750a5bd321567b4c0352403101bf292124c953d 1852 1851 2024-09-25T03:05:00Z Ben 2 wikitext text/x-wiki The [https://mjbots.com/products/moteus-n1 moteus-n1] is a motor controller from MJBots. The project is open-source, with firmware available [https://github.com/mjbots/moteus here]. {{infobox actuator |name = moteus-n1 |manufacturer = MJBots |cost = |purchase_link = https://mjbots.com/products/moteus-n1 |nominal_torque = |peak_torque = |weight = |dimensions = |gear_ratio = |voltage = 10 to 54VDC |cad_link = https://mjbots.com/products/moteus-n1 |interface = CAN-FD |gear_type = }} [[Category: Actuators]] cf862aff52cea254fc3f87b9a58e1bad359c6be1 0 435 1861 2024-10-15T00:46:07Z Ben 2 Created page with "Anydrive actuators, developed by ETH Zurich's Robotic Systems Lab, are high-performance, series elastic actuators designed for robotics. They offer precise torque, position, a..." wikitext text/x-wiki Anydrive actuators, developed by ETH Zurich's Robotic Systems Lab, are high-performance, series elastic actuators designed for robotics. They offer precise torque, position, and impedance control, making them suitable for legged robots like ANYmal. These actuators integrate motors, custom springs, high-precision encoders, and power electronics in a compact, water-resistant design. Anydrive's modular nature simplifies assembly and maintenance, and they communicate over CAN and CANopen protocols with ROS compatibility.<ref>https://rsl.ethz.ch/robots-media/actuators/anydrive.html</ref><ref>https://rsl.ethz.ch/robots-media/actuators.html</ref><ref>https://researchfeatures.com/anymal-unique-quadruped-robot-conquering-harsh-environments/</ref> == References == <references /> [[Category:Actuators]] 27d09c960fadbe2463e8c3e05a654c2e0ed9443c 1862 1861 2024-10-15T00:51:36Z Ben 2 wikitext text/x-wiki '''ANYdrive''' is a high-performance actuator developed by [[ANYbotics]], a Swiss robotics company specializing in mobile legged robots. ANYdrive actuators are designed for robust and versatile robotic applications, particularly in challenging environments. ==Overview== ANYdrive combines electric motor technology with integrated sensors and control electronics to provide dynamic and precise motion control. The actuator is used in ANYbotics' quadrupedal robots, such as the [[ANYmal]] robot. ==Features== * '''High torque density''': ANYdrive actuators deliver high torque in a compact form factor. * '''Integrated sensors''': Equipped with position, velocity, torque, and temperature sensors. * '''Robust design''': Suitable for harsh environments with dust and water protection. * '''Control electronics''': Integrated controllers for precise motion control. ==Applications== ANYdrive is utilized in various robotic platforms for research, inspection, and industrial automation. ==References== <references> <ref name="ANYbotics ANYdrive">{{cite web | title=ANYdrive Actuator | url=https://www.anybotics.com/anydrive/ | publisher=ANYbotics | accessdate=October 14, 2024}}</ref> </references> 921a85dd0bd23a7161e318ed06a482b6a4625c11 1863 1862 2024-10-15T00:53:03Z Ben 2 wikitext text/x-wiki '''ANYdrive''' is a high-performance actuator developed by ANYbotics, a Swiss robotics company specializing in mobile legged robots.<ref name="ANYbotics ANYdrive">{{cite web | title=ANYdrive Actuator | url=https://www.anybotics.com/anydrive/ | publisher=ANYbotics | accessdate=October 14, 2024}}</ref> ANYdrive actuators are designed for robust and versatile robotic applications, particularly in challenging environments. ==Overview== ANYdrive combines electric motor technology with integrated sensors and control electronics to provide dynamic and precise motion control.<ref name="ANYbotics ANYdrive" /> The actuator is used in ANYbotics' quadrupedal robots, such as the ANYmal robot. ==Features== * '''High torque density''': ANYdrive actuators deliver high torque in a compact form factor.<ref name="ANYbotics ANYdrive" /> * '''Integrated sensors''': Equipped with position, velocity, torque, and temperature sensors. * '''Robust design''': Suitable for harsh environments with dust and water protection. * '''Control electronics''': Integrated controllers for precise motion control. ==Applications== ANYdrive is utilized in various robotic platforms for research, inspection, and industrial automation.<ref name="ANYbotics ANYdrive" /> ==References== <references /> 884275d4bba4eb380945931cc35aa2b771d0a994 1864 1863 2024-10-15T00:53:39Z Ben 2 wikitext text/x-wiki '''ANYdrive''' is a high-performance actuator developed by ANYbotics, a Swiss robotics company specializing in mobile legged robots.<ref name="ANYbotics ANYdrive"><ref>https://www.anybotics.com/anydrive/</ref> ANYdrive actuators are designed for robust and versatile robotic applications, particularly in challenging environments. ==Overview== ANYdrive combines electric motor technology with integrated sensors and control electronics to provide dynamic and precise motion control.<ref name="ANYbotics ANYdrive" /> The actuator is used in ANYbotics' quadrupedal robots, such as the ANYmal robot. ==Features== * '''High torque density''': ANYdrive actuators deliver high torque in a compact form factor.<ref name="ANYbotics ANYdrive" /> * '''Integrated sensors''': Equipped with position, velocity, torque, and temperature sensors. * '''Robust design''': Suitable for harsh environments with dust and water protection. * '''Control electronics''': Integrated controllers for precise motion control. ==Applications== ANYdrive is utilized in various robotic platforms for research, inspection, and industrial automation.<ref name="ANYbotics ANYdrive" /> ==References== <references /> 1632ae12ee3f63d4ec842ecb74457141d06c89c5 1865 1864 2024-10-15T00:53:51Z Ben 2 wikitext text/x-wiki '''ANYdrive''' is a high-performance actuator developed by ANYbotics, a Swiss robotics company specializing in mobile legged robots.<ref><ref>https://www.anybotics.com/anydrive/</ref> ANYdrive actuators are designed for robust and versatile robotic applications, particularly in challenging environments. ==Overview== ANYdrive combines electric motor technology with integrated sensors and control electronics to provide dynamic and precise motion control.<ref name="ANYbotics ANYdrive" /> The actuator is used in ANYbotics' quadrupedal robots, such as the ANYmal robot. ==Features== * '''High torque density''': ANYdrive actuators deliver high torque in a compact form factor.<ref name="ANYbotics ANYdrive" /> * '''Integrated sensors''': Equipped with position, velocity, torque, and temperature sensors. * '''Robust design''': Suitable for harsh environments with dust and water protection. * '''Control electronics''': Integrated controllers for precise motion control. ==Applications== ANYdrive is utilized in various robotic platforms for research, inspection, and industrial automation.<ref name="ANYbotics ANYdrive" /> ==References== <references /> f30b77f049ef6846c5f6d73a916316c0aa39bcb7 1866 1865 2024-10-15T00:54:29Z Ben 2 wikitext text/x-wiki '''ANYdrive''' is a high-performance actuator developed by ANYbotics, a Swiss robotics company specializing in mobile legged robots.<ref name="ANYbotics ANYdrive ">https://www.anybotics.com/anydrive/</ref> ANYdrive actuators are designed for robust and versatile robotic applications, particularly in challenging environments. ==Overview== ANYdrive combines electric motor technology with integrated sensors and control electronics to provide dynamic and precise motion control.<ref name="ANYbotics ANYdrive" /> The actuator is used in ANYbotics' quadrupedal robots, such as the ANYmal robot. ==Features== * '''High torque density''': ANYdrive actuators deliver high torque in a compact form factor.<ref name="ANYbotics ANYdrive" /> * '''Integrated sensors''': Equipped with position, velocity, torque, and temperature sensors. * '''Robust design''': Suitable for harsh environments with dust and water protection. * '''Control electronics''': Integrated controllers for precise motion control. ==Applications== ANYdrive is utilized in various robotic platforms for research, inspection, and industrial automation.<ref name="ANYbotics ANYdrive" /> ==References== <references /> 2a631698de9978dcc12b527452340e7621a55edc 0 436 1867 2024-10-15T00:57:09Z Ben 2 Created page with "'''HEBI Robotics''' is an American company that specializes in modular robotic components and systems. Founded in 2014 and based in Pittsburgh, Pennsylvania, HEBI Robotics pro..." wikitext text/x-wiki '''HEBI Robotics''' is an American company that specializes in modular robotic components and systems. Founded in 2014 and based in Pittsburgh, Pennsylvania, HEBI Robotics provides hardware and software tools to create custom robots quickly and affordably.<ref name="HEBI About">https://www.hebirobotics.com/about</ref> ==Products== HEBI offers modular actuators, known as '''X-Series Actuators''', which integrate motors, gearboxes, encoders, and control electronics into compact modules.<ref name="HEBI Products">https://www.hebirobotics.com/products</ref> ==Features== * '''Modularity''': Allows for rapid prototyping and custom configurations. * '''Integrated control''': Built-in controllers enable precise motion control. * '''Software tools''': Provides APIs and software development kits for robot control. ==Applications== HEBI's products are used in research, education, and industry for applications such as manipulation, mobile robotics, and human-robot interaction.<ref name="HEBI Applications">https://www.hebirobotics.com/applications</ref> ==References== <references /> c9bc1465a61acce4433a2b862c3fae973d3cf8dd 1868 1867 2024-10-15T00:57:30Z Ben 2 wikitext text/x-wiki '''HEBI Robotics''' is an American company that specializes in modular robotic components and systems. Founded in 2014 and based in Pittsburgh, Pennsylvania, HEBI Robotics provides hardware and software tools to create custom robots quickly and affordably.<ref name="HEBI About">https://www.hebirobotics.com/about</ref> ==Products== HEBI offers modular actuators, known as '''X-Series Actuators''', which integrate motors, gearboxes, encoders, and control electronics into compact modules. ==Features== * '''Modularity''': Allows for rapid prototyping and custom configurations. * '''Integrated control''': Built-in controllers enable precise motion control. * '''Software tools''': Provides APIs and software development kits for robot control. ==Applications== HEBI's products are used in research, education, and industry for applications such as manipulation, mobile robotics, and human-robot interaction.<ref name="HEBI Applications">https://www.hebirobotics.com/applications</ref> ==References== <references /> 39a37dc77dbca9036d948ba74a921c9ce2fcb482 1869 1868 2024-10-15T00:57:53Z Ben 2 wikitext text/x-wiki '''HEBI Robotics''' is an American company that specializes in modular robotic components and systems. Founded in 2014 and based in Pittsburgh, Pennsylvania, HEBI Robotics provides hardware and software tools to create custom robots quickly and affordably.<ref name="HEBI About">https://www.hebirobotics.com/about</ref> ==Products== HEBI offers modular actuators, known as '''X-Series Actuators''', which integrate motors, gearboxes, encoders, and control electronics into compact modules.<ref>https://www.hebirobotics.com/actuators</ref> ==Features== * '''Modularity''': Allows for rapid prototyping and custom configurations. * '''Integrated control''': Built-in controllers enable precise motion control. * '''Software tools''': Provides APIs and software development kits for robot control. ==Applications== HEBI's products are used in research, education, and industry for applications such as manipulation, mobile robotics, and human-robot interaction.<ref name="HEBI Applications">https://www.hebirobotics.com/applications</ref> ==References== <references /> 98f415a6947c08f3ab1cd206b93d92d08c92b8d0 0 437 1870 2024-10-15T00:58:10Z Ben 2 Created page with "'''CubeMars''' is a manufacturer of robotic actuators, motors, and related components. The company focuses on providing high-performance products for robotics, drones, and aut..." wikitext text/x-wiki '''CubeMars''' is a manufacturer of robotic actuators, motors, and related components. The company focuses on providing high-performance products for robotics, drones, and automation industries.<ref name="CubeMars About">https://www.cubemars.com/about-us</ref> ==Products== * '''Brushless DC Motors''': High-efficiency motors suitable for robotics.<ref name="CubeMars Products">https://www.cubemars.com/products</ref> * '''Actuators''': Compact actuators with integrated control electronics. * '''Robot Joints''': Modular joints for robotic arms and manipulators. ==Features== * '''High Torque''': Motors and actuators designed for high-torque applications. * '''Compact Design''': Products are designed to be space-efficient. * '''Integrated Solutions''': Offers components that integrate multiple functions. ==Applications== CubeMars products are used in robotic arms, exoskeletons, drones, and other automation equipment.<ref name="CubeMars Applications">https://www.cubemars.com/applications</ref> ==References== <references /> ae0a5abb146997b5cdb86a719f667d9a23f7f5e5 0 438 1871 2024-10-15T00:58:20Z Ben 2 Created page with "'''mjbots''' is a robotics company that develops open-source hardware and software for high-performance robotic systems. Founded by Ben Katz, the company provides components s..." wikitext text/x-wiki '''mjbots''' is a robotics company that develops open-source hardware and software for high-performance robotic systems. Founded by Ben Katz, the company provides components such as actuators, controllers, and sensors.<ref name="mjbots About">https://mjbots.com/pages/about</ref> ==Products== * '''qdd100 Servo Actuator''': A high-torque direct-drive actuator.<ref name="mjbots Products">https://mjbots.com/collections/all</ref> * '''Moteus Controller''': An open-source motor controller for high-bandwidth control. * '''Components''': Various mechanical and electronic components for robotics. ==Features== * '''Open Source''': Hardware and software designs are open-source. * '''High Performance''': Components designed for dynamic and responsive control. * '''Community Support''': Active community contributing to development. ==Applications== mjbots products are used in legged robots, robotic arms, and other advanced robotic systems.<ref name="mjbots Applications">https://mjbots.com/applications</ref> ==References== <references /> a9228b33a46d270fb0894791bf088cb64426b611 1874 1871 2024-10-15T00:58:53Z Ben 2 wikitext text/x-wiki '''mjbots''' is a robotics company that develops open-source hardware and software for high-performance robotic systems. Founded by Ben Katz, the company provides components such as actuators, controllers, and sensors.<ref name="mjbots About">https://mjbots.com/pages/about</ref> ==Products== * '''qdd100 Servo Actuator''': A high-torque direct-drive actuator.<ref name="mjbots Products">https://mjbots.com/collections/all</ref> * '''Moteus Controller''': An open-source motor controller for high-bandwidth control. * '''Components''': Various mechanical and electronic components for robotics. ==Features== * '''Open Source''': Hardware and software designs are open-source. * '''High Performance''': Components designed for dynamic and responsive control. * '''Community Support''': Active community contributing to development. ==Applications== mjbots products are used in legged robots, robotic arms, and other advanced robotic systems.<ref name="mjbots Applications">https://mjbots.com/applications</ref> ==References== <references /> [[Category:Actuators]] [[Category:Open Source]] dafd1c8f81778549f9c6187d891185cd452f87ee 0 439 1872 2024-10-15T00:58:28Z Ben 2 Created page with "'''DYNAMIXEL''' is a line of smart actuators developed by the Korean robotics company Robotis. DYNAMIXEL servos are widely used in robotics for their integrated design, which..." wikitext text/x-wiki '''DYNAMIXEL''' is a line of smart actuators developed by the Korean robotics company Robotis. DYNAMIXEL servos are widely used in robotics for their integrated design, which combines motor, gearbox, controller, and networking capabilities.<ref name="Robotis DYNAMIXEL">https://www.robotis.us/dynamixel/</ref> ==Features== * '''Integrated Design''': Combines motor, reduction gear, controller, and network functionality. * '''Daisy-Chain Networking''': Multiple units can be connected in series for simplified wiring. * '''Programmable''': Supports various control modes and parameters. * '''Feedback''': Provides real-time feedback on position, speed, load, voltage, and temperature. ==Product Lines== * '''AX Series''': Entry-level actuators for education and hobbyists. * '''MX Series''': Mid-level actuators with improved precision and control. * '''X-Series''': Advanced actuators with high performance and reliability. * '''Pro Series''': Professional-grade actuators for industrial applications. ==Applications== DYNAMIXEL servos are used in humanoid robots, robotic arms, animatronics, and research projects.<ref name="Robotis Applications">https://www.robotis.us/applications/</ref> ==References== <references /> 2fdbba6f7732fe93e91325df2b89cd9d68b07aab 0 440 1873 2024-10-15T00:58:36Z Ben 2 Created page with "'''Lin Engineering''' is an American company specializing in the design and manufacturing of stepper motors and motion control solutions. Founded in 1987 and headquartered in..." wikitext text/x-wiki '''Lin Engineering''' is an American company specializing in the design and manufacturing of stepper motors and motion control solutions. Founded in 1987 and headquartered in Morgan Hill, California, Lin Engineering serves various industries including medical, automotive, and industrial automation.<ref name="Lin Engineering About">https://www.linengineering.com/about-us/</ref> ==Products== * '''Stepper Motors''': High-precision motors ranging from standard to custom designs.<ref name="Lin Engineering Products">https://www.linengineering.com/products/</ref> * '''Integrated Motors''': Motors with built-in controllers and drivers. * '''Motion Controllers''': Standalone controllers for precise motion applications. * '''Accessories''': Gearboxes, encoders, and other motion control components. ==Features== * '''Customization''': Offers tailored solutions to meet specific application needs. * '''High Precision''': Motors designed for accurate positioning and repeatability. * '''Quality Assurance''': Adheres to strict quality control standards. ==Applications== Lin Engineering's products are used in medical devices, robotics, aerospace, and other precision-driven fields.<ref name="Lin Engineering Applications">https://www.linengineering.com/applications/</ref> ==References== <references /> 4598e3505b4ff49ec7ab65dd78e39e231a7e97ce 0 441 1876 2024-10-15T20:10:50Z Ben 2 Created page with "[https://sima.ai/ SiMa AI] is a manufacturer of RISC-based neural network inference boards." wikitext text/x-wiki [https://sima.ai/ SiMa AI] is a manufacturer of RISC-based neural network inference boards. fbb95138933c47669934129706fa578a8a8404ff 0 442 1877 2024-10-15T20:11:31Z Ben 2 Created page with "The [https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-nano/product-development/ Jetson Nano] and [https://www.nvidia.com/en-us/autonomous-machines/embe..." wikitext text/x-wiki The [https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-nano/product-development/ Jetson Nano] and [https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/jetson-orin/ Jetson Orin] are embedded devices produced by Nvidia for edge neural network inference. c0240e802b3315a5ad00b5a0b5519b289f353807 0 64 1888 285 2024-10-24T20:38:01Z Ben 2 Ben moved page [[Agibot]] to [[AGIBot]] wikitext text/x-wiki The "Agi" in their name stands for "Artificial General Intelligence". Their CEO gave a detailed presentation here: https://www.youtube.com/watch?v=ZwjxbDVbGpU&t=1471s {{infobox company | name = Agibot | country = China | website_link = https://www.agibot.com/ | robots = [[RAISE-A1]] }} [[Category:Companies]] 73984c3a371c72de33880735a346e5eb1c3be56b 1890 1888 2024-10-24T20:44:19Z Ben 2 wikitext text/x-wiki The "AGI" in their name stands for "Artificial General Intelligence". {{infobox company | name = Agibot | country = China | website_link = https://www.agibot.com/ | robots = [[RAISE-A1]] }} == Notable People == * [https://www.youtube.com/@%E7%A8%9A%E6%99%96%E5%90%9B 稚晖君] == Links == * [https://www.youtube.com/watch?v=ZwjxbDVbGpU&t=1471s Presentation] * [https://github.com/AgibotTech/agibot_x1_hardware AGIBot X1 Hardware] * [https://github.com/AgibotTech/agibot_x1_train AGIBot X1 Training Code] * [https://github.com/AgibotTech/agibot_x1_infer AGIBot X1 Inference Code] [[Category:Companies]] b8e52493948169a340e97f1c2a226db054656e2c 1891 1890 2024-10-24T20:45:23Z Ben 2 /* Links */ wikitext text/x-wiki The "AGI" in their name stands for "Artificial General Intelligence". {{infobox company | name = Agibot | country = China | website_link = https://www.agibot.com/ | robots = [[RAISE-A1]] }} == Notable People == * [https://www.youtube.com/@%E7%A8%9A%E6%99%96%E5%90%9B 稚晖君] == Links == * [https://www.agibot.com/ Website] * [https://www.youtube.com/watch?v=ZwjxbDVbGpU&t=1471s Presentation] * [https://github.com/AgibotTech/agibot_x1_hardware AGIBot X1 Hardware] * [https://github.com/AgibotTech/agibot_x1_train AGIBot X1 Training Code] * [https://github.com/AgibotTech/agibot_x1_infer AGIBot X1 Inference Code] [[Category:Companies]] f527436f5e30aa2b737b5f356ebb8bc2eacea210 0 443 1889 2024-10-24T20:38:01Z Ben 2 Ben moved page [[Agibot]] to [[AGIBot]] wikitext text/x-wiki #REDIRECT [[AGIBot]] a3dd58b9e61331e93f00fff769e504f7ba29f8c4 0 121 1892 434 2024-10-25T07:18:38Z Ben 2 wikitext text/x-wiki Figure 01 is a humanoid robot from [[Figure AI]]. {{infobox robot | name = Figure 01 | organization = [[Figure AI]] | height = 167.6 cm | weight = 60 kg | payload = 20 kg | runtime = 5 Hrs | speed = 1.2 M/s | video_link = https://www.youtube.com/watch?v=48qL8Jt39Vs | cost = }} [[Category:Stompy, Expand!]] [[Category:Robots]] 3e845df3b9f5a8e86ca31d18cb831bd78c69d235 0 122 1893 435 2024-10-25T07:18:47Z Ben 2 wikitext text/x-wiki Figure AI is building a humanoid robot called [[Figure 01]]. {{infobox company | name = Figure AI | country = USA | website_link = https://www.figure.ai/ | robots = [[Figure 01]] }} [[Category:Stompy, Expand!]] [[Category:Companies]] ba66fbc3e02a573b155b3bd19f79b79652a722c6 1894 1893 2024-10-29T05:34:00Z Ben 2 wikitext text/x-wiki '''Figure Inc.''' is a robotics company focused on developing autonomous humanoid robots for industrial and potential consumer applications. Founded in 2022 in Silicon Valley, the company aims to address labor shortages by deploying general-purpose robots capable of performing complex tasks. Its current model, Figure 02, incorporates NVIDIA’s AI computing technologies, including NVIDIA H100 GPUs, enabling high-performance vision, dexterity, and conversational AI. {{infobox company | name = Figure AI | country = USA | website_link = https://www.figure.ai/ | robots = [[Figure 01]] }} == History == Founded by Brett Adcock, Figure demonstrated rapid early progress in humanoid robotics, leveraging synthetic data from NVIDIA Isaac Sim. The company’s humanoid robots were first deployed in testing environments, such as BMW's production line, to demonstrate their real-world applications. == Technology and Partnerships == Figure’s robots utilize NVIDIA’s robotics stack, with models trained on NVIDIA DGX for large-scale AI capabilities and designed using the NVIDIA Omniverse platform. Despite having limited in-house AI expertise, Figure's collaboration with OpenAI has furthered its AI models’ conversational abilities, marking a unique blend of perception and interaction capabilities. Figure's Series B funding announcement, amounting to $675 million, supports continued research and deployment of these robotic solutions in industrial settings. == Goals and Future Directions == Figure aims to commercialize its robots for industries such as manufacturing and logistics, with a potential consumer rollout on the horizon. Their humanoid models are designed for adaptability across different environments, driven by evolving AI, perception, and mobility technologies. [[Category:Companies]] 90248f2b4c9b99a1f4a8473a47fd0c6fd227a1de 0 444 1896 2024-11-01T09:49:09Z Kris 97 Created page with "== About == Humanoid is the first AI and robotics company in the UK creating the world’s leading, commercially scalable, and safe humanoid robots. Humanoid's founder is Arte..." wikitext text/x-wiki == About == Humanoid is the first AI and robotics company in the UK creating the world’s leading, commercially scalable, and safe humanoid robots. Humanoid's founder is Artem Sokolov, investor and serial entrepreneur. == Description == The company’s general-purpose humanoid robots are developed to be flexible and capable of adapting to complex environments. They will expand the market for industrial automation, including use cases in manufacturing, warehouses and logistics, among others. 0fd50b8499c06b40dce751d58ef90ced5b76eccf 0 444 1897 1896 2024-11-01T09:54:48Z Kris 97 wikitext text/x-wiki == About == [https://thehumanoid.ai/ Humanoid] is the first AI and robotics company in the UK creating the world’s leading, commercially scalable, and safe humanoid robots.<br> Humanoid's founder is Artem Sokolov, investor and serial entrepreneur. == Description == The company’s general-purpose humanoid robots are developed to be flexible and capable of adapting to complex environments. They will expand the market for industrial automation, including use cases in manufacturing, warehouses and logistics, among others. 3b4f98f5a7ae90ddf30798477c10373665f481eb 0 121 1898 1892 2024-11-09T19:38:54Z Ben 2 update figure article wikitext text/x-wiki Figure 01 is a humanoid robot developed by [[Figure AI]], a robotics company founded in 2022 by Brett Adcock. {{infobox robot | name = Figure 01 | organization = [[Figure AI]] | height = 167.6 cm | weight = 60 kg | payload = 20 kg | runtime = 5 Hrs | speed = 1.2 M/s | video_link = https://www.youtube.com/watch?v=48qL8Jt39Vs | cost = Unknown }} The robot is designed to perform a variety of tasks across industries such as manufacturing, logistics, warehousing, and retail. Standing at 1.6 meters tall and weighing 60 kilograms, Figure 01 is fully electric and boasts a runtime of approximately five hours on a single charge. It is equipped with advanced artificial intelligence systems, including integrations with OpenAI's models, enabling it to process and reason from language, as well as engage in real-time conversations with humans. === Development and Features === The development of Figure 01 progressed rapidly, with the company unveiling the prototype in October 2023, less than a year after its founding. The robot achieved dynamic walking capabilities shortly thereafter. In early 2024, Figure AI announced a partnership with OpenAI to enhance the robot's AI capabilities, allowing it to perform tasks such as making coffee by observing human actions and engaging in conversations. === Applications === Figure 01 is intended for deployment in various sectors to address labor shortages and perform tasks that are unsafe or undesirable for humans. In January 2024, Figure AI entered into an agreement with BMW to utilize the robot in automotive manufacturing facilities. Demonstrations have showcased Figure 01 performing tasks like assembling car parts and interacting with humans in real-time. === Limitations === Despite its advancements, Figure 01 has several limitations: * '''Task Complexity''': The robot is currently capable of performing structured and repetitive tasks but may struggle with complex or unstructured environments that require advanced problem-solving skills. * '''Speed and Efficiency''': Demonstrations have shown that Figure 01 operates at a slower pace compared to human workers, which could impact its efficiency in time-sensitive applications. * '''Physical Capabilities''': While designed to mimic human dexterity, the robot's ability to handle delicate or intricate objects is still under development, limiting its applicability in tasks requiring fine motor skills. * '''Battery Life''': With a runtime of approximately five hours, Figure 01 requires regular recharging, which could hinder continuous operation in certain industrial settings. * '''Environmental Adaptability''': The robot's performance in varying environmental conditions, such as extreme temperatures or uneven terrains, has not been extensively tested, potentially limiting its deployment in diverse settings. These limitations highlight the ongoing challenges in developing general-purpose humanoid robots capable of seamlessly integrating into human-centric environments. [[Category:Stompy, Expand!]] [[Category:Robots]] a4072e8c1663d04388c97c5fe1d60a6949707c3e 0 122 1899 1894 2024-11-09T19:39:37Z Ben 2 add figure 02 wikitext text/x-wiki '''Figure Inc.''' is a robotics company focused on developing autonomous humanoid robots for industrial and potential consumer applications. Founded in 2022 in Silicon Valley, the company aims to address labor shortages by deploying general-purpose robots capable of performing complex tasks. Its current model, Figure 02, incorporates NVIDIA’s AI computing technologies, including NVIDIA H100 GPUs, enabling high-performance vision, dexterity, and conversational AI. {{infobox company | name = Figure AI | country = USA | website_link = https://www.figure.ai/ | robots = [[Figure 01]], [[Figure 02]] }} == History == Founded by Brett Adcock, Figure demonstrated rapid early progress in humanoid robotics, leveraging synthetic data from NVIDIA Isaac Sim. The company’s humanoid robots were first deployed in testing environments, such as BMW's production line, to demonstrate their real-world applications. == Technology and Partnerships == Figure’s robots utilize NVIDIA’s robotics stack, with models trained on NVIDIA DGX for large-scale AI capabilities and designed using the NVIDIA Omniverse platform. Despite having limited in-house AI expertise, Figure's collaboration with OpenAI has furthered its AI models’ conversational abilities, marking a unique blend of perception and interaction capabilities. Figure's Series B funding announcement, amounting to $675 million, supports continued research and deployment of these robotic solutions in industrial settings. == Goals and Future Directions == Figure aims to commercialize its robots for industries such as manufacturing and logistics, with a potential consumer rollout on the horizon. Their humanoid models are designed for adaptability across different environments, driven by evolving AI, perception, and mobility technologies. [[Category:Companies]] 83415b894632ea121e13b80e8d11fd5b0c2ef934 1904 1899 2024-11-09T19:49:33Z Ben 2 wikitext text/x-wiki '''Figure Inc.''' is a robotics company focused on developing autonomous humanoid robots for industrial and potential consumer applications. Founded in 2022 in Silicon Valley, the company aims to address labor shortages by deploying general-purpose robots capable of performing complex tasks. Its current model, [[Figure 02]], incorporates NVIDIA’s AI computing technologies, including NVIDIA H100 GPUs, enabling high-performance vision, dexterity, and conversational AI. {{infobox company | name = Figure AI | country = USA | website_link = https://www.figure.ai/ | robots = [[Figure 01]], [[Figure 02]] }} == History == Founded by Brett Adcock, Figure demonstrated rapid early progress in humanoid robotics, leveraging synthetic data from NVIDIA Isaac Sim. The company’s humanoid robots were first deployed in testing environments, such as BMW's production line, to demonstrate their real-world applications. == Technology and Partnerships == Figure’s robots utilize NVIDIA’s robotics stack, with models trained on NVIDIA DGX for large-scale AI capabilities and designed using the NVIDIA Omniverse platform. Despite having limited in-house AI expertise, Figure's collaboration with OpenAI has furthered its AI models’ conversational abilities, marking a unique blend of perception and interaction capabilities. Figure's Series B funding announcement, amounting to $675 million, supports continued research and deployment of these robotic solutions in industrial settings. == Goals and Future Directions == Figure aims to commercialize its robots for industries such as manufacturing and logistics, with a potential consumer rollout on the horizon. Their humanoid models are designed for adaptability across different environments, driven by evolving AI, perception, and mobility technologies. [[Category:Companies]] d328b03ed32178ba87676942afaefd3a667fd19f 0 1 1900 1895 2024-11-09T19:40:05Z Ben 2 /* Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable mw-collapsible mw-collapsed" |+ Resources |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable mw-collapsible mw-collapsed" |+ Actuators |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[AGIBot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]], [[Figure 02]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Humanoid]] | |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} 29633789ff1e0a0a81464521b58a49bdeb699940 1905 1900 2024-11-15T20:05:27Z Ben 2 /* Robots */ wikitext text/x-wiki Welcome to the Humanoid Robots wiki! This is a free resource to learn about humanoid robots. As you're looking around, if something feels incomplete or out-of-date, please update it so that we can continue providing high-quality information to the community. Feel free to join our [https://discord.gg/kscale Discord community] for more real-time discussion! === Getting Started === [[Getting Started with Humanoid Robots]] {| class="wikitable mw-collapsible mw-collapsed" |+ Resources |- ! Name ! Comments |- | [[Underactuated Robotics]] | High-quality open-source course from MIT |- | [https://www.youtube.com/watch?v=LiNgr1tz49I&list=PLZnJoM76RM6ItAfZIxJYNKdaR_BobleLY Advanced Robot Dynamics] | High-quality open-source course from CMU |- | [https://www.youtube.com/watch?v=6rUdAOCNXAU&list=PLZnJoM76RM6KugDT9sw5zhAmqKnGeoLRa Optimal Control] | High-quality open-source course from CMU |- | [https://www.epfl.ch/labs/lasa/mit-press-book-learning/#chapter1 Learning for Adaptive and Reactive Robot Control] | Textbook for graduate-level courses in robotics |- | [[Learning algorithms]] | Resources related with training humanoid models in simulation and real environments |- | [[Servo Design]] | A reference for servos that you can use |- | [[:Category:Guides]] | Category for pages which act as guides |- | [[:Category:Courses]] | Category for pages about useful courses related to robotics |- | [[:Category:Electronics]] | Category for pages about electronics topics |- | [[:Category:Hardware]] | Category for pages relating to hardware |- | [[:Category:Firmware]] | Category for pages relating to firmware |- | [[:Category:Software]] | Category for pages relating to software |- | [[:Category:Teleop]] | Category for pages relating to teleoperation |- | [[:Category:Papers]] | Category for humanoid robotics papers |- | [[:Category:Non-humanoid Robots]] | Category for pages relating to non-humanoid robots |- | [[Contributing]] | How to contribute to the wiki |} === Components === {| class="wikitable mw-collapsible mw-collapsed" |+ Actuators |- | [[SPIN Servo]] |- | [[DEEP Robotics J60]] |- | [[Robstride]] |- | [[MyActuator]] |- | [[Encos]] |- | [[Steadywin]] |- | [[Elemental Motors]] |- | [[Anydrive]] |- | [[HEBI]] |- | [[ZeroErr]] |- | [[CubeMars]] |- | [[MjBots]] |- | [[Dynamixel]] |- | [[LinEngineering]] |- | [[Faradyi]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Inference Boards |- | [[Jetson]] |- | [[SiMi AI]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Motor Controllers |- | [[VESCular6]] |- | [[ODrive]] |- | [[OBot]] |- | [[Solo Motor Controller]] |- | [[moteus-n1]] |- | [[K-Scale Motor Controller]] |} {| class="wikitable mw-collapsible mw-collapsed" |+ Communication Protocols |- | [[Controller Area Network (CAN)]] |- | [[Inter-Integrated Circuit (I2C)]] |- | [[Serial Peripheral Interface (SPI)]] |- | [[EtherCAT]] |} === Robots === {| class="wikitable" |- ! Company ! Robots |- | [[1X]] | [[Eve]], [[Neo]] |- | [[AGIBot]] | [[RAISE-A1]] |- | [[Agility]] | [[Cassie]], [[Digit]] |- | [[Anthrobotics]] | [[Anthro]] |- | [[Apptronik]] | [[Valkyrie]], [[Draco]], [[QDH]], [[Apollo]] |- | [[AstriBot Corporation]] | [[Astribot S1]] |- | [[Beijing Humanoid Robot Innovation Center]] | [[Tiangong]] |- | [[Boardwalk Robotics]] | [[Nadia]], [[Alex]] |- | [[Booster Robotics]] | [[BR002]] |- | [[Boston Dynamics]] | [[HD Atlas]], [[Atlas]] |- | [[Bot Company]] | |- | [[DATAA Robotics]] | [[XR4]] |- | [[Deep Robotics]] | [[Wukong-IV]] |- | [[MagicLab,_DREAME]] | [[MagicBot]] |- | [[Engineered Arts]] | [[Ameca]] |- | [[FDROBOT]] | [[T1]] |- | [[Figure AI]] | [[Figure 01]], [[Figure 02]] |- | [[Fourier Intelligence]] | [[GR-1]] |- | [[GALBOT]] | [[GALBOT]] |- | [[Haier]] | [[Kuavo (Kuafu)]] |- | [[Honda Robotics]] | [[ASIMO]] |- | [[Humanoid]] | |- | [[Hyperspawn Robotics]] | [[Shadow-1]] |- | [[Instituto Italiano]] | [[iCub]] |- | [[Kawasaki Robotics]] | [[Kaleido]], [[Friends]] |- | [[Kayra.org]] | [[Kayra]] |- | [[Kepler]] | [[K1]] |- | [[K-Scale Labs]] | [[Stompy]] |- | [[Kind Humanoid]] | [[Mona]] |- | [[LASER Robotics]] | [[HECTOR V2]] |- | [[LEJUROBOT]] | [[Kuavo]] |- | [[LimX Dynamics]] | [[CL-1]] |- | [[MenteeBot]] | [[MenteeBot (Robot)]] |- | [[Mirsee Robotics]] | [[Beomni]], [[Mirsee]] |- | [[NASA]] | [[Valkyrie]], [[Robonaut2]] |- | [[NEURA Robotics]] | [[4NE-1]] |- | [[Noetix]] | [[Dora]] |- | [[PAL Robotics]] | [[Kangaroo]], [[REEM-C]], [[TALOS]] |- | [[PaXini]] | [[Tora]] |- | [[POINTBLANK]] | [[DROPBEAR]] |- | [[Pollen Robotics]] | [[Reachy]] |- | [[Proxy]] | |- | [[Rainbow Robotics]] | [[HUBO]] |- | [[Robotera]] | [[XBot]], [[Starbot]] |- | [[Sanctuary]] | [[Phoenix]] |- | [[SoftBank Robotics]] | [[Pepper]], [[NAO]] |- | [[Stanford Robotics Lab]] | [[OceanOneK]] |- | [[SuperDroid Robots]] | [[Rocky]] |- | [[SUPCON]] | [[Navigator α]] |- | [[System Technology Works]] | [[ZEUS2Q]] |- | [[Tesla]] | [[Optimus]] |- | [[THK]] | |- | [[Toyota Research Institute]] | [[Punyo]], [[T-HR3]] |- | [[UBTech]] | [[Walker X]], [[Panda Robot]], [[Walker S]] |- | [[UC Berkeley]] | [[Berkeley Blue]] |- | [[Unitree]] | [[H1]], [[G1]] |- | [[University of Tehran]] | [[Surena IV]] |- | [[Westwood Robotics]] | [[THEMIS]] |- | [[WorkFar]] | [[WorkFar Syntro]] |- | [[Xiaomi]] | [[CyberOne]] |- | [[Xpeng]] | [[PX5]] |} ca5e68b1d8eca95335639f7d11f143984b4c6d58 0 445 1901 2024-11-09T19:45:32Z Ben 2 Created page with "Figure 02 is a second-generation humanoid robot developed by Figure AI, an AI robotics company based in Sunnyvale, California. Unveiled in August 2024, Figure 02 represents a..." wikitext text/x-wiki Figure 02 is a second-generation humanoid robot developed by Figure AI, an AI robotics company based in Sunnyvale, California. Unveiled in August 2024, Figure 02 represents a significant advancement in humanoid robotics, combining human-like dexterity with advanced artificial intelligence to perform a wide range of tasks across various industries.<ref name="prnewswire">{{cite web |title=Figure Unveils Figure 02, Its Second-Generation Humanoid, Setting New Standards in AI and Robotics |url=https://www.prnewswire.com/news-releases/figure-unveils-figure-02-its-second-generation-humanoid-setting-new-standards-in-ai-and-robotics-301894889.html |website=PR Newswire |date=August 15, 2024}}</ref> == Development and Features == Building upon its predecessor, Figure 01, the development of Figure 02 involved a comprehensive hardware and software redesign. Key features of Figure 02 include: * '''Speech-to-Speech Interaction''': Equipped with onboard microphones and speakers, Figure 02 can engage in natural speech-to-speech conversations with humans, utilizing custom AI models developed in collaboration with OpenAI.<ref name="prnewswire" /> * '''Vision-Language Model (VLM)''': The robot incorporates an onboard VLM that enables rapid common-sense visual reasoning, allowing it to interpret and respond to visual inputs effectively.<ref name="prnewswire" /> * '''Enhanced Battery Capacity''': A custom 2.25 kWh battery pack housed in the torso provides over 50% more energy compared to the previous generation, extending operational runtime.<ref name="prnewswire" /> * '''Integrated Wiring''': The design features integrated cabling for power and communication, resulting in concealed wires, increased reliability, and a more streamlined appearance.<ref name="prnewswire" /> * '''Advanced Vision System''': Six onboard RGB cameras power an AI-driven vision system, enabling the robot to perceive and understand its environment with high accuracy.<ref name="prnewswire" /> * '''Fourth-Generation Hands''': The robot's hands are equipped with 16 degrees of freedom and human-equivalent strength, facilitating the execution of a wide range of human-like tasks.<ref name="prnewswire" /> * '''Increased Computational Power''': With triple the computational and AI inference capabilities of its predecessor, Figure 02 can perform real-world AI tasks fully autonomously.<ref name="prnewswire" /> == Applications == Figure 02 is designed for deployment across various sectors, including manufacturing, logistics, warehousing, and retail. In recent tests at BMW Manufacturing's facility in Spartanburg, South Carolina, the robot successfully performed AI data collection and use case training, demonstrating its potential to enhance productivity and efficiency in industrial settings.<ref name="prnewswire" /> == Limitations == Despite its advancements, Figure 02 has certain limitations: * '''Environmental Adaptability''': The robot's performance in diverse and unstructured environments requires further validation to ensure reliability across various settings. * '''Task Complexity''': While capable of performing a wide range of tasks, Figure 02 may encounter challenges with highly complex or nuanced activities that demand advanced problem-solving skills. * '''Battery Life''': Although the battery capacity has been enhanced, the operational runtime may still be insufficient for continuous use in certain industrial applications without periodic recharging. * '''Physical Capabilities''': The robot's ability to handle extremely delicate or intricate objects is still under development, which may limit its applicability in tasks requiring fine motor skills. Addressing these limitations is essential for the broader adoption and integration of Figure 02 into various industries. == See Also == * [https://www.figure.ai/ Figure AI] * [https://openai.com/ OpenAI] * [https://en.wikipedia.org/wiki/Humanoid_robot Humanoid Robot] == References == <references /> == External Links == * [https://www.figure.ai/ Figure AI Official Website] * [https://www.youtube.com/@figureai Figure AI YouTube Channel] 2d7999f719284d7c7d8ad33cba91a3e7bba56c69 1902 1901 2024-11-09T19:47:45Z Ben 2 wikitext text/x-wiki Figure 02 is a second-generation humanoid robot developed by Figure AI, an AI robotics company based in Sunnyvale, California. {{infobox robot | name = Figure 02 | organization = Figure AI | video = [https://www.youtube.com/watch?v=7bFsxW_8ABE Video] | cost = Not publicly disclosed | height = 5 ft 6 in (167.6 cm) | weight = 155 lbs (70 kg) | speed = 1.2 m/s | lift force = 44 lbs (20 kg) | battery life = 5 hours | battery capacity = 2.25 kWh | purchase = Not available for purchase | number made = Limited prototypes | dof = 16 in hands | status = In development }} Unveiled in August 2024, Figure 02 represents a significant advancement in humanoid robotics, combining human-like dexterity with advanced artificial intelligence to perform a wide range of tasks across various industries.<ref name="prnewswire">{{cite web |title=Figure Unveils Figure 02, Its Second-Generation Humanoid, Setting New Standards in AI and Robotics |url=https://www.prnewswire.com/news-releases/figure-unveils-figure-02-its-second-generation-humanoid-setting-new-standards-in-ai-and-robotics-301894889.html |website=PR Newswire |date=August 15, 2024}}</ref> == Development and Features == Building upon its predecessor, Figure 01, the development of Figure 02 involved a comprehensive hardware and software redesign. Key features of Figure 02 include: * '''Speech-to-Speech Interaction''': Equipped with onboard microphones and speakers, Figure 02 can engage in natural speech-to-speech conversations with humans, utilizing custom AI models developed in collaboration with OpenAI.<ref name="prnewswire" /> * '''Vision-Language Model (VLM)''': The robot incorporates an onboard VLM that enables rapid common-sense visual reasoning, allowing it to interpret and respond to visual inputs effectively.<ref name="prnewswire" /> * '''Enhanced Battery Capacity''': A custom 2.25 kWh battery pack housed in the torso provides over 50% more energy compared to the previous generation, extending operational runtime.<ref name="prnewswire" /> * '''Integrated Wiring''': The design features integrated cabling for power and communication, resulting in concealed wires, increased reliability, and a more streamlined appearance.<ref name="prnewswire" /> * '''Advanced Vision System''': Six onboard RGB cameras power an AI-driven vision system, enabling the robot to perceive and understand its environment with high accuracy.<ref name="prnewswire" /> * '''Fourth-Generation Hands''': The robot's hands are equipped with 16 degrees of freedom and human-equivalent strength, facilitating the execution of a wide range of human-like tasks.<ref name="prnewswire" /> * '''Increased Computational Power''': With triple the computational and AI inference capabilities of its predecessor, Figure 02 can perform real-world AI tasks fully autonomously.<ref name="prnewswire" /> == Applications == Figure 02 is designed for deployment across various sectors, including manufacturing, logistics, warehousing, and retail. In recent tests at BMW Manufacturing's facility in Spartanburg, South Carolina, the robot successfully performed AI data collection and use case training, demonstrating its potential to enhance productivity and efficiency in industrial settings.<ref name="prnewswire" /> == Limitations == Despite its advancements, Figure 02 has certain limitations: * '''Environmental Adaptability''': The robot's performance in diverse and unstructured environments requires further validation to ensure reliability across various settings. * '''Task Complexity''': While capable of performing a wide range of tasks, Figure 02 may encounter challenges with highly complex or nuanced activities that demand advanced problem-solving skills. * '''Battery Life''': Although the battery capacity has been enhanced, the operational runtime may still be insufficient for continuous use in certain industrial applications without periodic recharging. * '''Physical Capabilities''': The robot's ability to handle extremely delicate or intricate objects is still under development, which may limit its applicability in tasks requiring fine motor skills. Addressing these limitations is essential for the broader adoption and integration of Figure 02 into various industries. == See Also == * [https://www.figure.ai/ Figure AI] * [https://openai.com/ OpenAI] * [https://en.wikipedia.org/wiki/Humanoid_robot Humanoid Robot] == References == <references /> == External Links == * [https://www.figure.ai/ Figure AI Official Website] * [https://www.youtube.com/@figureai Figure AI YouTube Channel] cf26614e9f70fe5646d489b9082a024a7cdf8c77 1903 1902 2024-11-09T19:49:08Z Ben 2 wikitext text/x-wiki Figure 02 is a second-generation humanoid robot developed by [[Figure AI]], an AI robotics company based in Sunnyvale, California. {{infobox robot | name = Figure 02 | organization = Figure AI | video = [https://www.youtube.com/watch?v=7bFsxW_8ABE Video] | cost = Not publicly disclosed | height = 5 ft 6 in (167.6 cm) | weight = 155 lbs (70 kg) | speed = 1.2 m/s | lift force = 44 lbs (20 kg) | battery life = 5 hours | battery capacity = 2.25 kWh | purchase = Not available for purchase | number made = Limited prototypes | dof = 16 in hands | status = In development }} Unveiled in August 2024, Figure 02 represents a significant advancement in humanoid robotics, combining human-like dexterity with advanced artificial intelligence to perform a wide range of tasks across various industries.<ref name="prnewswire">{{cite web |title=Figure Unveils Figure 02, Its Second-Generation Humanoid, Setting New Standards in AI and Robotics |url=https://www.prnewswire.com/news-releases/figure-unveils-figure-02-its-second-generation-humanoid-setting-new-standards-in-ai-and-robotics-301894889.html |website=PR Newswire |date=August 15, 2024}}</ref> == Development and Features == Building upon its predecessor, Figure 01, the development of Figure 02 involved a comprehensive hardware and software redesign. Key features of Figure 02 include: * '''Speech-to-Speech Interaction''': Equipped with onboard microphones and speakers, Figure 02 can engage in natural speech-to-speech conversations with humans, utilizing custom AI models developed in collaboration with OpenAI.<ref name="prnewswire" /> * '''Vision-Language Model (VLM)''': The robot incorporates an onboard VLM that enables rapid common-sense visual reasoning, allowing it to interpret and respond to visual inputs effectively.<ref name="prnewswire" /> * '''Enhanced Battery Capacity''': A custom 2.25 kWh battery pack housed in the torso provides over 50% more energy compared to the previous generation, extending operational runtime.<ref name="prnewswire" /> * '''Integrated Wiring''': The design features integrated cabling for power and communication, resulting in concealed wires, increased reliability, and a more streamlined appearance.<ref name="prnewswire" /> * '''Advanced Vision System''': Six onboard RGB cameras power an AI-driven vision system, enabling the robot to perceive and understand its environment with high accuracy.<ref name="prnewswire" /> * '''Fourth-Generation Hands''': The robot's hands are equipped with 16 degrees of freedom and human-equivalent strength, facilitating the execution of a wide range of human-like tasks.<ref name="prnewswire" /> * '''Increased Computational Power''': With triple the computational and AI inference capabilities of its predecessor, Figure 02 can perform real-world AI tasks fully autonomously.<ref name="prnewswire" /> == Applications == Figure 02 is designed for deployment across various sectors, including manufacturing, logistics, warehousing, and retail. In recent tests at BMW Manufacturing's facility in Spartanburg, South Carolina, the robot successfully performed AI data collection and use case training, demonstrating its potential to enhance productivity and efficiency in industrial settings.<ref name="prnewswire" /> == Limitations == Despite its advancements, Figure 02 has certain limitations: * '''Environmental Adaptability''': The robot's performance in diverse and unstructured environments requires further validation to ensure reliability across various settings. * '''Task Complexity''': While capable of performing a wide range of tasks, Figure 02 may encounter challenges with highly complex or nuanced activities that demand advanced problem-solving skills. * '''Battery Life''': Although the battery capacity has been enhanced, the operational runtime may still be insufficient for continuous use in certain industrial applications without periodic recharging. * '''Physical Capabilities''': The robot's ability to handle extremely delicate or intricate objects is still under development, which may limit its applicability in tasks requiring fine motor skills. Addressing these limitations is essential for the broader adoption and integration of Figure 02 into various industries. == See Also == * [https://www.figure.ai/ Figure AI] * [https://openai.com/ OpenAI] * [https://en.wikipedia.org/wiki/Humanoid_robot Humanoid Robot] == References == <references /> == External Links == * [https://www.figure.ai/ Figure AI Official Website] * [https://www.youtube.com/@figureai Figure AI YouTube Channel] 325f68987e2859e083fe8889383c623c44c52fdd 0 446 1906 2024-11-15T20:05:43Z Ben 2 Created page with "[https://www.bot.co/ Bot Company] is a company founded in 2024 by Kyle Vogt, the former founder of Cruise." wikitext text/x-wiki [https://www.bot.co/ Bot Company] is a company founded in 2024 by Kyle Vogt, the former founder of Cruise. 6bf8161b4ddf273617feca757bf142a8e9e2faaf 6 447 1907 2024-11-22T06:20:53Z Ben 2 wikitext text/x-wiki Aesthetic A b291e15610c6e6c3833d01bc1d8352383708c5ee 6 448 1908 2024-11-22T06:21:08Z Ben 2 wikitext text/x-wiki Aesthetic B 41e50c3d7ec3c36ca5fc0d0171305cd222e7ed67 6 449 1909 2024-11-22T06:22:03Z Ben 2 wikitext text/x-wiki Aesthetic C b0cdc5dd34f47f3f0ce9a1aa791139758ad57e3d 6 450 1910 2024-11-22T06:22:16Z Ben 2 wikitext text/x-wiki Aesthetic D f32dc1454233b3bec074d5ecac3a660185a6db63 6 451 1911 2024-11-22T06:22:31Z Ben 2 wikitext text/x-wiki Aesthetic E 57b101b6d460c80b5fe7789b2b065ffebb7d8af8 6 452 1912 2024-11-22T06:22:45Z Ben 2 wikitext text/x-wiki Aesthetic F e34cf1b551eedc0b7e3675ec47d4c6f755595d7c 0 5 1913 842 2024-11-22T06:25:24Z Ben 2 /* Logos */ wikitext text/x-wiki [[File:Logo.png|right|200px|thumb|The K-Scale Labs logo.]] [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} == Logos == Here are some other versions of the K-Scale Labs logo. <gallery> Logo.png K-Scale Labs Raw Logo.png K-Scale Padded Logo.png K-Scale Raw Padded Logo.png KScale Raw Transparent Padded Logo.png K-Scale Raw Padded White Logo.png Aesthetic A.png Aesthetic B.png Aesthetic C.png Aesthetic D.png Aesthetic E.png Aesthetic F.png </gallery> === Adding Padding === Here's a helpful command to add padding to an image using ImageMagick: <syntaxhighlight lang="bash"> convert \ -channel RGB \ -negate \ -background black \ -alpha remove \ -alpha off \ -gravity center \ -scale 384x384 \ -extent 512x512 \ logo.png \ logo_padded.png </syntaxhighlight> [[Category:Companies]] [[Category:K-Scale]] cfaaf2cbaee558d22c2cd70c62082bbc3bb87051 1915 1913 2024-11-23T22:56:07Z Ben 2 wikitext text/x-wiki [[File:Logo.png|right|200px|thumb|The K-Scale Labs logo.]] [https://kscale.dev/ K-Scale Labs] is building an open-source humanoid robot called [[Stompy]]. {{infobox company | name = K-Scale Labs | country = United States | website_link = https://kscale.dev/ | robots = [[Stompy]] }} == Logos == Here are some other versions of the K-Scale Labs logo. <gallery> Logo.png K-Scale Labs Raw Logo.png K-Scale Padded Logo.png K-Scale Raw Padded Logo.png KScale Raw Transparent Padded Logo.png K-Scale Raw Padded White Logo.png Aesthetic A.png Aesthetic B.png Aesthetic C.png Aesthetic D.png Aesthetic E.png Aesthetic F.png Aesthetic G.png </gallery> === Adding Padding === Here's a helpful command to add padding to an image using ImageMagick: <syntaxhighlight lang="bash"> convert \ -channel RGB \ -negate \ -background black \ -alpha remove \ -alpha off \ -gravity center \ -scale 384x384 \ -extent 512x512 \ logo.png \ logo_padded.png </syntaxhighlight> [[Category:Companies]] [[Category:K-Scale]] cabcaf24ca61b4c89ac2d4e8ce3dfad66eaeaa18 6 453 1914 2024-11-23T22:53:59Z Ben 2 wikitext text/x-wiki Aesthetic G 015616930e92eeff7fc9982c21c00c4bbf8c1b7a